ICRU REPORT 59

Clinical Proton Dosimetry

Part I: Beam Production,
Beam Delivery and
Measurement of
Absorbed Dose

issued: 15 December 1998

INTERNATIONAL COMMISSION ON RADIATION

UNITS AND MEASUREMENTS
7910 WOODMONT AVENUE
BETHESDA, MARYLAND 20814
U.S.A.
 THE INTERNATIONAL COMMISSION
ON RADIATION UNITS AND MEASUREMENTS

INDIVIDUALS PARTICIPATING IN THE PREPARATION OF THIS REPORT

Commission Membership during Report Committee

Preparation of this Report L. VERHEY, Chairman
A. ALLISY,Chairman University of California San Fransico
A. WAMBERSIE, Vice Chairman San Francisco, California, U.S.A.
R. S . CASWELL,
Secretary H.BLA~TMAN
P. M. DELUCA,JR Paul Scherrer Institute
K. Do1 Villigen,
- Switzerland
P. M. DELUCA
University of Wisconsin Medical School
Madison, Wisconsin, U.S.A.
D. MILLER
Loma Linda University Medical Center
Lorna Linda, California, U S A .
L. S. TAYLOR,
Honorary Chairman and Consultants to the Report Committee
Member Emeritus
P. ANDREO
Current Commission Membership International Atomic Energy Agency
Vienna,Austria
A. WAMBERSIE, Chairman
P. M. DELUCA, JR,Vice Chairman H. BICHSEL
R. S. CASWELL, Secretary Seattle, Washington, U.SA.
K. Doi D. JONES
L. Feinendegen National Accelerator Centre
M. Inokuti Faure, South Africa
H. Menzel S. WNCKIER
H. Paretzke Universite' Catholique de Louvain
G. F. Whitmore Bruxelles, Belgium
L. S. Taylor, Honorary Chairman and
Member Emeritus Commission Sponsors
H. D. Wyckoff, Honorary Chairman H. H. Ross1
A. Allisy, Honorary Chairman Columbia University
Principal Scientific Counselor New York, New York, U.S.A.
A. WAMBERSIE
Universite' Catholique 0% Louvain
Assistant Secretary Bruxelles, Belgium
Managing Editor of ZCRUNEWS

Counselor to the Commission on Financial Affairs

0.W. LINTON

The Commission wishes to express its appreciation to the individuals involved in the preparation of this Report for
the time and effort which they devoted to this task and to express its appreciation to the organizations with which
they are affiliated.
Copyright B International Commission on Radiation Units and Measurements 1998
(For detailed information ofthe availability of this and other ICRU Reports, see page 54.)

Preface
Scope of ICRU Activities
The International Commission on Radiation Units
and Measurements (ICRU), since its inception in
1925, has had as its principal objective the develop-
ment of internationally acceptable recommendations
regarding:
1. Quantities and units of radiation and radio-
activity,
2. Procedures suitable for the measurement and
application of these quantities in clinical radiology
and radiobiology and
3. Physical data needed in the application of these
procedures, the use of which tends to assure unifor-
mity in reporting.
The Commission also considers and makes similar
types of recommendations for the radiation protec-
tion field. In this connection, its work is carried out
in close cooperation with the International Commis-
sion on Radiological Protection (ICRP).
Policy
The ICRU endeavors to collect and evaluate the
latest data and information pertinent to the prob-
lems of radiation measurement and dosimetry and to
recommend the most acceptable values and tech-
niques for current use.
The Commission's recommendations are kept un-
der continual review in order to keep abreast of the
rapidly expanding uses of radiation.
The ICRU feels that it is the responsibility of
national organizations to introduce their own de-
tailed technical procedures for the development and
maintenance of standards. However, it urges that all
countries adhere as closely as possible to the interna-
tionally recommended basic concepts of radiation
quantities and units.
The Commission feels that its responsibility lies in
developing a system of quantities and units having
the widest possible range of applicability. Situations
may arise from time to time when an expedient
solution of a current problem may seem advisable.
Generally speaking, however, the Commission feels
that action based on expediency is inadvisable from a
long-term viewpoint; it endeavors to base its decision
on the long-range advantages to be expected.
The ICRU invites and welcomes constructive com-
ments and suggestions regarding its recommenda-
tions and reports. These may be transmitted to the
chairman.
"
U
E
t
a
Current Program
The Commission has divided its field of interest
into 12 technical areas and has assigned
- one or more
members of the Commission the responsibility for
identification of potential topics for new ICRU activi-
ties in each area. The technical areas are:
Quantities and Units
Radiation Therapy
Diagnostic Radiology
Nuclear Medicine
Radiobiology
Radioactivity
Radiation Physics- X Rays, Gamma Rays and Electrons
Radiation Physics- Neutrons and Heavy Particles
Radiation Protection
Radiation Chemistry
Nuclear Data
Theoretical Aspects
The Commission is currently considering the possibil-
ity of expan- its program to encompass nonionizing
radiation, particularly the quantities and units asp&.
The actual preparation of ICRU reports is carried
out by ICRU report committees. One or more Com-
mission members serve as sponsors to each commit-
tee and provide close liaison with the Commission.
The currently active report committees are:
Absorbed Dose Standards for Photon Irradiation and Their
Dissemination
Assessments of Image Quality in Nuclear Medicine
Beta Rays for Therapeutic Applications
Bone Densitometry
Chest Radiography - Assessment of Image Quality
Clinical Proton Dosimetry-Part 11: Dose Specification for
Reporting, Treatment Planning and Radiation Quality
Determination of Body Burdens for Radionuclides
Dose and Volume Specification for Reporting Intracavtiary
Therapy in Gynecology
Dose Specificationin Nuclear Medicine
~osirn&ricProcedures in Diagnostic Radiology
Fundamental Quantities and Units
~ a m m o ~ i - a ~ h ; Assessment
- of Image Quality
Measurement of Operational Quantities for Neutrons
Nuclear Data for Neutron and Proton Radiotherapy and for
Radiation Protection
Prescribing, Recording and Reporting Electron Beam Therapy
Requirements for Radioecological Sampling
Retrospective Assessment of Exposure to Ionizing Radiation
ROC Analysis
Requirements for Radioecological Sampling
Stopping Power for Heavy Ions
Tissue Substitutes, Characteristics of Biological Tissue and tD
Phantoms for Ultrasound
a?
3
oi
m
ICRU's Relationships with Other 2a
Organizations 2
3
In addition to its close relationship with the ICRP,
the ICRU has developed relationships with other B
organizations interested in the problems of radiation
quantities, units and measurements. Since 1955, the
ICRU has had an official relationship with the World
Health Organization (WHO), whereby the ICRU is iii
looked to for primary guidance in matters of radia- National Cancer Institute of the U.S. Department of
tion units and measurements and, in turn, the WHO Health and Human Services and the ~nternational
assists in the world-wide dissemination of the Com- Atomic Energy Agency.
mission's recommendations. In 1960, the ICRU en- In addition, during the last 10 years, financial
tered into consultative status with the International support has been received from the following organi-
Atomic Energy Agency. The Commission has a for- zations:
mal relationship with the United Nations Scientific American Society for Therapeutic Radiology and Oncology
Committee on the Effects of Atomic Radiation Atomic Energy Control Board
(UNSCEAR),whereby ICRU observers are invited to Bayer AG
sttend UNSCEAR meetings. The Commission and Central Electricity Generating Board
Commissariat a 1'EnergieAtomique
the International Organization for Standardization Dutch Society for Radiodiagnostics
(ISO) informally exchange notifications of meetings, Eastman Kodak Company
and the ICRU is formally designated for liaison with Ebara Corporation
two of the I S 0 technical committees. The ICRU also Electricit6 de France
Fuji Medical Systems
corresponds and exchanges final reports with the General Electric Company
following organizations: Hitachi, Ltd.
Bureau International de MBtrologie LQgale International Radiation Protection Association
Bureau International des Poids e t Mesures International Society of Radiology
Council for International Organizations of Medical Sciences Italian RadiologicalAssociation
European Commission Japan Industries Association of Radiological Systems
Food and Agriculture Organization of the United Nations Konica Corporation
International Committee of Photobiology National Electrical Manufacturers Association
International Council of Scientific Unions Philips Medical Systems, Incorporated
International Electrotechnical Commission Radiation Research Society
International Labor Office Scanditronix AB
International Organization for Medical Physics Siemens Aktiengesellschaft
International Radiation Protection Association Sumitomo Heavy Industries, Ltd.
International Union of Pure and Applied Physics Theratronics
United Nations Educational, Scientific and Cultural Organization Toshiba Corporation
The Commission has found its relationship with University Hospital Lund, Sweden
World Health Organization
all of these organizations fruitful and of substantial
benefit to the ICRU program. Relations with these I n addition to the direct monetary support pro-
other international bodies do not affect the basic vided by these organizations, many organizations
m i a t i o n of the ICRU with the International Society provide indirect support for the Commission's pro-
of Radiology. gram. This support is provided in many forms,
including, among others, subsidies for (1) the time of
individuals participating in ICRU activities, (2) travel
Operating Funds
costs involved in ICRU meetings and (3) meeting
In the early days of its existence, the ICRU facilities and services.
operated essentially on a voluntary basis, with the In recognition of the fact that its work is made
travel and operating costs being borne by the parent possible by the generous support provided by all of
organization of the participants. (Only token assis- the organizations supporting its program, the Com-
tance was originally available from the Interna- mission expresses its deep appreciation.
tional Society of Radiology.) Recognizing the impract-
ibility of continuing this mode of operation on an Andre Wambersie
indefinite basis, operating funds were sought from Chairman, ICRU
various sources.
In recent years, principal financial support has Bruxelles, Belgium
been provided by the European Commission, the 15 October 1998
 Contents
...
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

.
1 Introduction ............................................... 1
1.1 Rationale for Proton Therapy ............................. 1
1.2 History of Proton Therapy ................................ 2
1.3 Scope of this Report ...................................... 4
1.4 Relation to Existing Reports .............................. 5
.
2 Production of Proton Beams for Therapeutic Applications ... 6
2.1 Accelerators ............................................. 6
2.1.1 Cyclotrons ........................................ 6
2.1.1.1 Synchrocyclotrons .......................... 6
2.1.1.2 Isochronous Cyclotrons ..................... 7
2.1.2 Synchrotrons ...................................... 7
2.1.3 Linear Accelerators ................................ 8
2.2 Typical Operating Parameters ............................ 8
2.3 Beam Shaping and Delivery .............................. 8
2.3.1 Passive Scattering Techniques ...................... 8
2.3.2 Wobbling and Beam Scanning....................... 9
2.3.3 Spot Scanning ..................................... 10
.
3 Pertinent Quantities and Units ............................ 11
3.1 Physical Quantities ...................................... 11
3.2 Radiometric Quantities .................................. 11
3.3 Interaction Coefficients................................... 11
3.4 Dosimetric Quantities .................................... 12
3.5 Stochastic Quantities .................................... 12
.
4 Proton Interactions with Matter ........................... 13
4.1 Electromagnetic and Nuclear Interactions ................. 13
4.2 Energy Deposition ....................................... 13
4.2.1 Average Energy Loss ............................... 13
4.2.2 Microdosimetric Concepts .......................... 14
4.3 Factors which M e c t Proton Beam Characteristics .......... 14
.
5 Determination of Proton Absorbed Dose in Reference
Conditions............................................... 15
5.1 General Considerations for Proton Dosimetry .............. 15
5.2 Fluence Measurements with a Faraday Cup ................ 16
5.3 Absorbed Dose Measurements with a Calorimeter .......... 17
5.4 Absorbed Dose Measurements with an Ionization Chamber . . 18
5.4.1 Introduction ....................................... 18
5.4.2 Conceptual Description of Absorbed Dose
Determination .................................. 18
5.4.3 Interpretation of Ionization Chamber Response ....... 19
5.4.3.1 Considerations of Stopping Power ........... 19
5.4.3.2 Considerations of w and W .................. 21
5.4.4 Determination of Absorbed Dose to Water ............ 26
5.4.4.1 Determinations Based on Fluence ........... 26
5.4.4.2 Determinations Based on Air Kerma
Calibrations ............................ 27
5.4.4.3 Determinations Based on Absolute Dose to
Water Calibrations ...................... 27
.
6 Beam Monitoring and Relative Dosimetry ................. 29
6.1 Beam Monitoring ........................................ 29
6.1.1 Detectors for Monitoring Beam Intensity ............. 29
6.1.1.1 Ionization Chambers for Beam Monitoring ... 29
 6.1.1.2 Other Beam Monitoring Detectors . . . . . . . . . . .
6.1.2 Special Considerations for Monitoring of Dynamic
Beam Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2 Dose Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2.1 Detectors for Dose Distribution Measurements . . . . . . .
6.2.1.1 Silicon Diodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2.1.2 Films . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2.1.3 Other Detectors for Relative Dosimetry . . . . . .
6.2.2 Determination of Dose Distributions . . . . . . . . . . . . . . . . .
6.2.2.1 Measurement of Beam Range and Depth Dose
Characteristics ..........................
6.2.2.2 Lateral Dose Uniformity and Beam
Penumbra ..............................
.
7 Recommendations for Determination of Absorbed Dose in
a Phantom ...............................................
7.1 General Recommendations ...............................
7.1.1 Reference Dosimeter ...............................
7.1.2 Phantom Material and Reference Depth .............
7.2 Determination of Proton Absorbed Dose to Water using a
Calibrated Ionization Chamber .........................
7.2.1 NK-basedCalibrated Ionization Chamber ............
7.2.2 Nwbased Calibrated Ionization Chamber ............
7.3 Determination of Proton Absorbed Dose to Water using a
Calorimeter...........................................
7.4 Determination of Proton Absorbed Dose to Water using a
Faraday Cup ..........................................
7.5 Numerical Values of Required Quantities ..................
7.6 Summary of Recommendations ...........................
Appendix A
Quality Assurance Example ...............................
Appendix B
Dosimetry Worksheet......................................
References ...................................................
ICRU Reports ................................................
Index .........................................................
 Clinical Proton
Dosimetry -Part I: Beam
Production, Beam Delivery and
Measurement of Absorbed Dose
1. Introduction
1.1 Rationale for Proton Therapy
Proton beam therapy is a technologically advanced
means of achieving extremely precise radiation dose
distributions, resulting in less dose to the normal
tissues surrounding the target volume than is typi-
cally achievable with conventional x-ray or electron
beam therapy. As a result, higher doses can be safely
delivered to target volumes which are adjacent, or
close to critical normal structures which are vulner-
able to radiation injury. Higher doses to the malig-
nant cell population should yield a n increased prob-
ability of local tumor control (Suit et al., 1990;
Thames et al., 1992). Achievement of local tumor
control is essential in cancer therapy and has impor-
tant practical consequences for the survival of cancer
patients.
To illustrate the importance of the problem, in
1980,1.2 million new cancer cases were diagnosed in
the countries of the European Union (Davis et al.,
1990; Doll, 1990; Muir and Boyle, 1990). Among
them, 65% presented with apparently localized tu-
mors. About two-thirds of these are cured by surgery,
radiation therapy or a combination of both tech-
niques. Unfortunately, local failure occurred in about
one-third of the patients who still had apparently
localized disease a t the first consultation (Devita,
1983; Devita and Korn, 1986).Even more pessimistic
data were reported by Silverberg et al. (Silverberg et
al., 1990) and Devita et al. (Devita et al., 1989).
Therefore, it can be expected that if local failure
could be reduced by 50%, cure rates could be im-
proved by 10-15% (Suit, 1982; Suit and Miralbell,
1989). Continued improvement in survival would be
expected to result from the implementation of sys-
tems of radiation treatment that provide better
physical dose localization, thereby permitting higher
dose to regions of disease while maintaining or
reducing complications resulting from the irradin-
tion of normal tissues.
As a consequence of their excellent physical dose
localization properties, protons are indicated for the
treatment of tumors which are close to critical
normal structures. In such cases, conventional irra-
diation techniques often cannot achieve adequate
tumor control due to the limitations imposed by
normal tissue tolerance.
Protons are also indicated for localized tumors
where other treatment modalities, such as surgery,
are judged likely to result in unacceptable patient
morbidity. Reduced normal tissue doses have the
potential of reducing the side effects of multimodal-
ity therapy when radiation, surgery andlor chemo-
therapy are used. Proton beam therapy should be
considered where preservation of organ function
could be compromised by alternative therapeutic
strategies (Suit and Urie, 1992; Wambersie et al.,
1992).
Protons traverse relatively straight paths through
a medium, slowing down continuously due to interac-
tions with electrons. A fraction of the protons d l
suffer nuclear interactions, in which they are effec-
tively removed from the beam. A depth dose distribu-
tion for a monoenergetic proton beam (Figure 1.1)
shows a region of dose1 rising slowly with depth,
called the "plateau7', followed by a dose maximum
called the "Bragg peak," with an amplitude three to
four times the entrance dose (Shipley et al., 1979).
Beyond the Bragg peak, dose rapidly approaches
zero. A superposition of Bragg peaks, created by
varying the energy (a technique known as energy
modulation) yields a region of relatively uniform
high dose, often called a spread-out Bragg peak
(SOBP), which can be designed to have a width
sufficient to cover the target volume. The SOBP dose
still falls rapidly towards zero beyond the distal
target edge, although modulation also has the effect
of increasing the entrance dose relative to an un-
modulated beam. Figure 1.1shows depth dose distri-
butions for a modulated and an unmodulated proton
beam of about 160 MeV, compared with a 10 MV
x-ray bremsstrahlung beam.
The term dose, in this Report, is used as an abbreviation of
absorbed dose.

Modulated Proton Beam
SOBP
Depth in Tissue, cm
Fig. 1.1. Depth-dose curves for a 10 MV bremsstrahlungx-ray
beam, an unmodulated 160 MeV proton beam (with a single
m w Bragg peak and for a moddated proton beam with a spread-
out Bragg peak (SOBP)width of about 7 cm (Shipleyet d.,
1979).
By varying the energy of the protons throughout
the beam cross-section, or by placing absorbers of
varying thickness between the beam source and a
patient, it is possible to conform the distal edge of the
high dose region to the distal edge of the target
volume. This technique is usually referred to as
energy- or range-modulation. In addition, dynamic
beam delivery systems may be employed to conform
the dose to the proximal edge of the target volume
(see Section 2.3). Beams from several directions may
be used to produce dose distributions that are confor-
mal with the 3-dimensional shape of the target
volume. Clinical results confirm the benefits of this
approach (Gragoudas et al., 1982; Suit et al., 1982a;
Suit et al., 1982b; Minakova et al., 1983; Austin-
Seymour et al., 1985; Munzenrider et al., 1985;
Saunders et al., 1985; Austin-Seymour et al., 1989;
Austin-Seymour et al., 1990).
The technology for development of proton treat-
ment facilities is readily available. The principles of
operation of proton accelerators capable of producing
beams with energies in the range of 250 MeV are
well understood and examples of such accelerators
are being used for proton radiation therapy (see
Table 1.1). Computer control systems provide the
simplicity of operation required for a hospital set-
ting. The technology of beam delivery systems, includ-
ing rotating gantries, is well developed. Three-
dimensional diagnostic imaging, particularly
computed tomography (CT) and magnetic resonance
imaging (MRI), is readily available and provides the
basis for tissue delineation and assignment of tissue
properties for proton beam treatment planning.
Three-dimensional treatment planning systems of-
fer capabilities for simulation and optimization of
patient treatments and the design of patient-specific
beam apertures and tissue compensators (Goitein
and Abrams, 1983; Goitein et al., 1983; Goitein and
Miller, 1983). Numerically controlled machine took
are available for fabricating patient-specific devices
from data generated by the treatment planning
system (Wagner, 1982). In addition, since normal
tissues are significantly spared by the dose distribu-
tions of proton beams, greater flexibility is available
for optimization of the treatment delivery schedule.
The relative biological effectiveness (RBE) of pro-
ton beams compared to 60Co gamma-rays is only
slightly greater than unity. Most of the RBE measure
ments for protons give values ranging from 1.0 to 1.2
relative to 60Co gamma-rays (Robertson et al., 1975;
Urano et al., 1984; Robertson et al., 1994; Gueulette
et al., 1996). This rather large range of values could
be due to differences in biological systems, end
points, proton beam characteristics, tissue depth.
dosimetry methods or experimental uncertainties.
No significant biological differences are expected
from such a small RBE, so the clinical experience
accumulated with x rays can be extrapolated to
protons, especially that related to normal tissue
tolerance (Suit et al., 1980).A weighting factor of 1.1
is typically applied to proton doses to account for
RBE effects, yielding a dose which should be biologi-
cally equivalent to x rays.
1.2 History of Proton Therapy
The use of proton beams for treatment of human
disease was first suggested by R. R. Wilson in 1946
(Wilson, 1946).At that time, Wilson recognized that
accelerators capable of generating proton be-=
with sufficient energy to provide a range in tissue
comparable to body dimensions, were under construc-
tion. Wilson noted that the mass of the proton would
cause it to travel in a nearly straight path through
tissue and that the energy deposition pattern of a
proton beam would produce high radiation doses
near the end of the range in a relatively narrow
region, now referred to as the Bragg peak (Figure
1.1). Wilson proposed the irradiation of localized
regions within the body with proton beams to pro-
vide maximum sparing of surrounding tissues. In his
1946 paper, Wilson proposed that rotating modula-
tor wheels could spread the Bragg peak over large
targets, transmission ionization chambers could be
employed to monitor patient dose and ionization
chambers could provide absolute dose calibration.
By 1954, C . A. Tobias and his associates a t the
University of California a t Berkeley had completed
animal irradiation studies and began small-field
treatment of the human pituitary gland with a 340
MeV proton beam from the 184 inch cyclotron (To-
bias et al., 1955; Tobias et al., 1958). Initially, cross-
fired beams penetrating through the entire head
were used, followed by the first application of Bragg
peak treatment. After 1957, these treatments were
converted to deuterons and alpha particles. A total of
274 patients had been treated by 1964 with 190 MeV
deuterons, 340 MeV protons or 900 MeV alpha
particles (Tobias et al., 1964). This project continued
to contribute knowledge, interest and technical ad-
vances essential to proton therapy (Castro et al.,
1985) until the shutdown of the cyclotron in 1987
followed by that of the Bevalac accelerator in 1992,
by which time approximately 2000 patients had been
treated with protons, deuterons and alpha particles.
Inspired by the work a t Berkeley, B. Larsson
began the development of proton therapy as a neuro-
surgical tool a t the Gustaf Werner Institute in
Uppsala, Sweden (Larsson et al., 1958; Larsson et
al., 1959; Leksell et al., 1960; Larsson, 1961; Larsson
et al., 1963; Larsson et al., 1974). Following the
neurosurgical work, the first proton treatments with
large fields, the first use of a scanned proton beam
delivery system and the first proton treatments with
a modulated Bragg peak were developed a t Uppsala
(GraEman et al., 1967; Larsson, 1967; Graffman and
Jung, 1970; G r d m a n e t al., 1975). By 1970, a total of
69 patients had received large-field, energy-modu-
lated proton treatments to disease sites in many
locations within the body (Graffman et al., 1985).
This facility resumed operation in 1988 following a
period of extensive equipment upgrade (Montelius et
al., 1991).
I n 1961, R. N. Kjellberg and associates (Kjellberg
et al., 1962; Kjellberg et al., 1983) began treatment of
human intracranial tumors a t the Harvard Cyclo-
tron Laboratory (HCL) using a 160 MeV proton
beam. Large-field fractionated proton therapy began
a t HCL in 1974 with a beam delivery system employ-
ing rotating range modulator wheels (Koehler et al.,
1975), passive scattering (Koehler et al., 1975), and
shaped range compensators (Wagner, 1982). Meth-
ods for three-dimensional treatment planning of
proton treatments were developed a t Massachusetts
General Hospital (Goitein and Abrams, 1983; Goit-
ein et al., 1983; Goitein and Miller, 1983). Dosimetry
methods were formulated (Verhey et al., 1979) and
techniques for patient positioning were developed
(Verhey et al., 1982). These systems allowed high-
precision, high-dose radiation therapy with proton
beams for many disease sites (Suit et al., 1975; Suit
et al., 1977; Suit et al., 1980; Suit et al., 1982b)
including proton treatment of ocular melanoma (Con-
stable and Koehler, 1974; Gragoudas et al., 1977). By
1991 over 5000 patients had been treated a t the
Harvard facility with extensive follow-up (Kliman et
al., 1984;Austin-Seymour et al., 1985; Munzenrider
et al., 1985) inspiring world-wide interest in the
development of hospital-based proton treatment ten-
ters.
oped in the former Soviet Union. The Joint Institute
for Nuclear Research (JINR) in Dubna began treat-
ing patients in 1968 with beam energies up to 200
MeV from a 680 MeV synchrocyclotron. This was
followed in 1969 by the initiation of proton therapy
at the Institute of Theoretical and Experimental
Physics (ITEP) in Moscow, where five beam energies
ranging from 70 MeV to 200 MeV can be selected
from a 10 GeV synchrotron. By 1987, 1360 patients
with disease in a variety of sites had been treated
(Minakova, 1987).The Leningrad Institute of Nuclear
Physics (LINP) a t Gatchina began treatment with a
1000 MeV proton beam in 1975. Intracranial treat-
ment of 508 patients using cross-fire irradiation with
small fields in a single fraction were reported (Raju
et al., 1987). Summaries of the former Soviet Union
proton treatment experience and facilities may be
found in several publications (Boone et al., 1977;
Riabukhin, 1982;Abrosimov et al., 1985; GraEman et
al., 1985).
Treatment with negative pions provided impor-
tant technical innovations that influenced the contin-
ued development of proton therapy. A pion treatment
facility a t The Los Alamos Scientific Laboratory in
the USA operated from 1974 to 1981 (Kligerman et
al., 1979) joined by the development of pion treat-
ments a t the Swiss Institute for Nuclear Research
(SIN) now known as the Paul Scherrer Institute
(PSI) in Villigen, Switzerland Wonessen et al., 1982)
and the TRIUMF laboratory in Vancouver, B.C.
Canada (Goodman et al., 1985).At PSI, multiple pion
beams were focused onto a single stationary spot and
the patient was moved within a cylinder of water
according to a planned pattern to deliver a dose
distribution that conformed to the defined target
volume. Developments from pion therapy that are
important for proton treatment include three-
dimensional conformal techniques for beam delivery,
patient immobilization and positioning methods and
the development of treatment planning tools for
conformal radiation therapy.
Two facilities for proton therapy were established
in Japan (Tsunemoto et al., 1985). In 1979 the
National Institute for Radiological Sciences at Chiba
began clinical trials for treatment of superficial
lesions with a 70 MeV beam. This facility was the
first to develop a spot scanning system for the
delivery of proton beams (Kanai et al., 1980). The
Proton Medical Research Center (PMRC)in Tsukuba
began treatment with a 250 MeV beam in 1983 and
by July, 1995 had treated 462 patients. The PMRC
facility produced a clinical proton beam by energy
degradation from 500 MeV. This facility was the first
to provide a vertical proton beam and the first to
make use of a multivane collimator for proton therapy
(Matsuda and Inamura, 1981).
 TABLE1.1-Proton therapy fmilities worldwide

Center Localion Proton energy (MeV) Number of patient8 treated Period'

LBLlLTCSF Berkeley, USA 740

Svedberg Lab. Uppsala, Sweden 185
MGWHCL Boston, USA 160
ITEP Moscow, Russia 200
JINR Dubna, Russia 680
JINR Dubna, Russia 680
LINP St. Petersburg (Gatchina), Russia 1000
NiXS Chiba, Japan 86
PMRC Tsukuba, Japan 250
PSI Villigen, Switzerland 72
Svedberg Lab. Uppsala, Sweden 60
Douglas Cyclotron Clatterbridge, UK 62
LLUMC Loma Linda, USA 250
UCL Louvain-la-Neuve, Belgium 85
CPO Orsay, France 73 (200)b
Centre A. Lacassagne Nice, France 65
NAC Faure, South Africa 200
Indiana Cyclotron Indianapolis, USA 200
UCSFNC Davis Davis, CA, USA 70
TOTAL
a Year of first treatment-year (month) when the patient numbers were tabulated or last year of operation, if facility is closed.
Although 636 patients were treated for uveal melanoma using a 73 MeV proton beam, since 1994, a 200 MeV beam has been available
and has been usedto treat 320 patients with brain tumors.

In Europe, a t the PSI in Villigen, Switzerland,
treatment of uveal melanoma was initiated in 1984
with a 72 MeV proton beam facility (the OPTIS
program) (Perret, 1989). This successful program
had treated 2,487 patients by the end of 1997.2Atthe
same institute, a new proton therapy program is
being developed, using scanned beams of up to 250
MeV delivered with an isocentric gantry (Pedroni et
al., 1989). The first patient treatments began in
1996.
At the Clatterbridge Hospital in the U.K., the
cyclotron initially designed for fast neutron therapy
has been used to treat uveal melanomas with 62
MeV protons since 1989. By June, 1998, 817 ocular
melanomas had been treated.2
Similarly, a t the Universite Catholique de Lou-
vain, in Louvain-la-Neuve, Belgium, where fast neu-
tron therapy has been ongoing since 1978, a feasibil-
ity study of proton therapy with 85 MeV protons was
initiated in 1991. Twenty-one patients were treated
by November, 1993. At that time, the installation
was dismantled and a new proton beam is now being
installed (Wambersie, 1995).
In Orsay, close to Paris, the synchrocyclotron
designed for physical experiments was adapted for
proton beam therapy at 73 MeV and was first used to
treat uveal melanomas in 1991. A total of 673
patients were treated through November, 1995. In
the same treatment room, the full 200 MeV energy is
now available for therapy and treatments of brain
Treatment numbers are from the Proton Therapy Cooperative
Group Newsletter of July, 1998 and were provided by Janet
Sisterson.
tumors were initiated in 1994 with a fixed, horizon-
tal beam.
In Nice, France, the hospital based cyclotron of the
Centre Antoine Lacassagne, initially designed for
neutron therapy, is now used for neutron and proton
therapy. Using 65 MeV protons, 1,010 uveal melano-
mas have been treated through 1997.
The first hospital-based center designed specifi-
cally for proton therapy was established by James
Slater a t Loma Linda University Medical Center in
California (Slater et al., 1988; Slater et al., 1991).
Treatment beams for four therapy rooms and a
research room are produced by a variable-energy 250
MeV synchrotron designed and built a t the Fermi
National Laboratory specifically for proton treat-
ment. The first patient was treated in October, 1990.
By 1991, proton beams varying from 100 MeV to 200
MeV were in clinical use on two horizontal beam
lines and the 200 MeV line was upgraded to 250 MeV
in 1992. The first isocentric beam delivery system
employing a rotating gantry was commissioned for
patient treatment a t Loma Linda in June, 1991 with
beam energies of 155, 200 and 250 MeV. Two addi-
tional gantry treatment rooms were commissioned
in 1994. The facility a t Loma Linda provides large-
field fractionated treatments, treatment of ocular
melanoma and stereotactic radiosurgical techniques
with a total capacity of approximately 100 patient
treatments per day. By July of 1998, over 3400
p a t i e ~ ~had
t s been treated in this facility.
Table 1.1lists proton therapy facilities around the
world, the number of patients treated, the proton
energy, the date of first treatment and the date for
which thi: treatment numbers are recorded. As of the
 middle of 1998, in excess of 21,500 patients had been
treated with protons worldwide, a number which is
growing a t a rate of approximately 2000 patients per
year (Sisterson, 1995). This increase in the number
of facilities and the growing clinical activity under-
scores the need for standardization of dosimetry.
1.3 Scope of this Report
The development of accurate and uniform stan-
dards for radiation treatment dosimetry has been a
continuing effort since the earliest days of radio-
therapy. Recommended techniques for the dosimetry
of photon and electron beams have becn developed
(NCRP, 1981; AAPM, 1983; IAEA, 1987) and con-
tinue to evolve. The expanding capabilities for treat-
ment with proton beams must be met with corre-
sponding efforts in standardizing accurate dosimetry
methods. This ICRU Report is intended to promote
uniformity of standards that will provide a basis for
world-wide comparison of clinical results and allow
the development of meaningful clinical trials.
This Report describes current practice in proton
therapy and recommends standards for the dosim-
etry of proton treatments. Established proton treat-
ment facilities might use this Report as a source of
information for the maintenance of accurate stan-
. dards. New facilities may build their procedures
from recommendations found in this Report and
planners of new facilities may examine alternatives
within current practice for the production and moni-
toring of treatment beams.
This Report will be published in two separate
parts. Part I includes a description of the interaction
of protons with matter, various methods of beam
production, the characteristics of proton beams in
and specific recommendations for dose calibration.
Part 11, to be published at a later time, will deal with
the influence of patient shape and tissue heterogene-
ity on dose distribution, a description of treatment
planning considerations, a definition of beam quality
in relation to microdosimetry and a discussion of the
clinical radiobiological effectiveness (RBE) of proton
beams relative to photon beams.
1.4 Relation to Existing Reports
The American Association of Physicists in Medi-
cine produced AAPM Report 16, "Protocol for Heavy
Charged Particle Therapy Beam Dosimetry,"in 1986
(AAPM, 1986). That report was prepared by Task
Group 20 of the AAPM Radiation Therapy Commit-
tee. Report 16 provides guidelines for the dosimetry
of therapy beams for heavy charged particles includ-
ing protons and heavier ions. Although the report
cites deficiencies in the basic data necessary to
provide a desired level of accuracy in dosimetry, it
has served as a valuable working document for
dosimetry standardization. The European Clinical
Heavy Particle Dosimetry Group (ECHED) pub-
lished a "Code of Practice for Clinical Proton Dosim-
etry" in 1991 (Vynckier et al., 1991) and a more
recent supplement (Vynckier et al., 1994). This code
of practice recommends techniques for determina-
tion of absorbed dose to tissue for clinical proton
beams.
This Report is more comprehensive than any
single document currently available. It is intended to
serve as a complete and self-contained reference for
determination of absorbed dose in proton beams as
used in clinical practice.
 2. Production of Proton Beams for Therapeutic Applications
2.1. Accelerators
Proton accelerators designed for therapy need to
produce beams with energy great enough to reach
the deepest tumors, usually considered to be in
excess of 200 MeV, a beam current adequate to
achic-~etreatment times comparable to those of
conventional x-ray treatment facilities for thc range
of field sizes and doses normally used in radio-
therapy -approximately 10 nA or more, and the
capability to provide uniform doses to the r m g e of
target volumes likely to be encountered. This corre-
sponds to a requirement of more than 5 x 101°
protons per second extracted from the accelerator
and an efficient beam spreading system.
Particle accelerators are normally built with ei-
ther a straight or a circular arrangement of the
components. In the circular, fixed orbit geometry
(synchrotron), each particle traverses the same path
repeatedly during the acceleration cycle, with a n
increasing magnetic field keeping the particles in a
fixed orbit. An extraction system is used to deflect
the particles outside of the accelerator. In the circu-
lar, variable radius geometry (cyclotron), groups of
particles, often referred to as 'bunches" are acceler-
ated in a fixed magnetic field and spiral outward as
they gain energy. In a linear accelerator (linac), each
particle generally traverses the acceleration cavity
only once. As a consequence, linear accelerators are
long, but do not require bending magnets to keep the
particles in orbit. In a linear accelerator, a large
number of bunches can be accelerated in a short time
span, thereby reducing the need for a large number
of protons per bunch which can be limited by space
charge considerations. Thus, linear accelerators can
usually accelerate higher currents than circular
machines. On the other hand, it is simpler to change
energies with some circular machines than with
linear machines. The selection of the best accelerator
for clinical proton therapy will be strongly influenced
by the weighting of the different characteristics such
as initial cost for the accelerator and the facility as a
whole, reliability, space availability, number of treat-
ment rooms, complexity and cost of operation, range
of energies desired, beam delivery technique, etc.
2.1.1 Cyclotrons
In a cyclotron, particles are accelerated by a high
voltage high frequency electric field in a gap between
two bending sections. Because both the magnetic
field and the acceleration frequency are constant,
cyclotrons are simple and inexpensive. As long as the
particles have a low energy and are therefore non-
relativistic, particles of different energies need the
same time for a full circle, so the machine can
simultaneously accelerate bunches of particles from
the lowest to the highest energy. Once the speed of
the particles has increased to a value where the
relativistic mass increase is substantial (of the order
of one percent), they get out of phase with the
accelerator field and further acceleration in this
simple way cannot be accomplished-that is, either
the magnetic field or the acceleration frequency
must be varied. This phenomenon occurs a t an
energy of approximately 10 MeV for protons, much
lower than the energy required for therapy. To
account for the increasing relativistic mass of the
particles, the frequency of the acceleration field can
be varied in synchrony with the speed of the particle.
This limits acceleration to a single bunch of protons
a t a time. An accelerator of this type is called a
synchrocyclotron or FM (frequency-modulated) cyclo-
tron.
Another way these relativistic effects have been
dealt with is by increasing the magnetic field with
radius, or by changing the shape of the pole faces to
increase the pathway of the protons in the magnetic
field, or both. All bunches of particles are simulta-
neously kept in an isochronous orbit by this accelera-
tor type which is, therefore, called a n isochronous
cyclotron.
In the following, the different accelerators are
discussed in two groups depending on whether they
use room temperature magnets or magnets cooled to
achieve the superconducting state.
2.1.1.1 Synchrocyclotrons. In the synchrocyclo-
tron, also called an FM (frequency modulated) cyclo-
tron, the increase in the relativistic mass of the
protons with energy is compensated by a decrease in
the frequency of the accelerating voltage, while the
magnetic field strength remains constant. This accel-
eration method requires pulsed operation, since one
proton bunch must exit the synchrocyclotron before
the acceleration of the next bunch begins. Synchrocy-
clotrons with conventional magnets are large in size
and mass. In principle, they can be tuned for differ-
ent energies, but currently operating machines of
this type have fixed energies. The extraction effi-
ciency of the synchrocyclotron is of the order of 90%
and extracted proton beam intensity is normally
more than adequate for therapy. The synchrocyclo-
tron is considered a very reliable accelerator, with no
stringent tolerances on the magnetic field shape.
This assertion is verified by the experience with the
Harvard syichrocyclotron which has operated for
proton therapy with a very high reliability since
1961 (Sisterson et al., 1991).
The superconducting synchrocyclotron has been
proposed as a low mass accelerator well-suited for
proton therapy (Blosser et al., 1991). The experience
with the superconducting type is somewhat limited,
the first machine having been installed in 1982.
Since superconducting cyclotrons can be much
smaller in diameter than conventional machines, the
mechanical tolerances are smaller, making the ion
source design, field shape tolerance a d frequency
modulation more critical.
2.1.1.2 Isochronous Cyclotrons. In contrast to
the synchrocyclotron, the isochronous cyclotron uses
a high voltage, constant frequency acceleration sys-
tem. An isochronous cyclotron compensates for the
increase in mass of the relativistic particle by increas-
ing the strength of the magnetic field traversed by
the particle with increasing radius of particle orbit.
The field shape required to achieve isochronisity
leads to axial defocusing in the magnetic field but
this is compensated by strong focusing hills and
valleys resulting in precisely defined geometric or-
bits and an intense beam. Since the frequency of the
accelerating field is constant, the isochronous cyclo-
tron can be operated with a very high pulse rate and
can be considered to be a continuous radiation
source.
Compared to a clinical synchrocyclotron, a fixed
energy isochronous cyclotron is heavier, of the order
of 80 t, and substantially more expensive. However,
much higher beam currents can be produced in the
isochronous cyclotron. Additional advantages of a
fixed energy machine are ease of operation, reliabil-
ity, and simplicity of extraction, making this design a
suitable solution for proton radiotherapy. A non-
superconducting isochronous cyclotron with high
field strength has been proposed (Beeckman et al.,
1990). Using a new magnet gap design, a reduced
magnet weight and reduced energy consumption are
predicted. Superconducting cyclotrons have clear
advantages over room temperature accelerators in
terms of size. Due to the higher magnetic fields
achievable, their diameters are approximately half
and the mass is reduced by an order of magnitude.
This is especially important if the accelerator is
mounted on a gantry as proposed by Blosser (Blosser
et al., 1991). They also consume much less energy.
Superconducting cyclotrons also have a common
disadvantage. If opening of the machine is necessary
for service or repair, longer down periods are the
result, since warming up and cooling down can take
a number of days. In addition, superconducting
equipment needs a liquid helium supply or a helium
liquifier. However, for fully tested accelerators in
routine operation, reopening may not be necessary
for long times. The development of magnetic reso-
nance imaging has demonstrated that superconduc-
tivity is a practical technology in the hospital environ-
ment. Operation of superconducting accelerators a t
research labs has demonstrated high levels of reliabil-
ity with infrequent need to bring the magnets to
room temperature.
2.1.2 Synchrotrons.
In a synchrotron, bending magnets keep a bunch
of protons in a fixed orbit during the acceleration
cycle. The frequency of the high voltage acceleration
system, installed in straight sections between the
bending magnets, is increased with increasing speed
of the particle and is closely coupled with an increase
in the magnetic field in the bending magnets. The
protons can be extracted a t any energy by either
single turn extraction or by slow extraction to achieve
longer pulses. Both room temperature and supercon-
ducting proton synchrotrons have been considered
as well as H- synchrotrons. Only room temperature
designs will be discussed below.
The first hospital-based proton radiotherapy facil-
ity installed a t Loma Linda University Medical
Center is based on a room temperature synchrotron
(Cole et al., 1987). A synchrotron has also been
proposed in Japan a t Tsukuba University for a new
hospital-based therapy facility (Fukurnoto et al.,
1989).
The synchrotron is a very flexible machine type in
terms of energy variation. With a synchrotron, i t is
feasible to use energy variation of the beam instead
of range shifting with a variable thickness absorber
to modulate the energy for depth control of the
proton beam. A possible limitation for this accelera-
tor is the maximum current per bunch which can be
captured. This is due to space charge effects which
depend on the injection energy. The injector at Lorna
Linda is a radio frequency quadrupole (RFQ) linac
(Kapchinskii and Teplyakov, 1970)of 2 MeV, while a t
Tsukuba, a 5 MeV RFQ linac is planned.
Synchrotrons with conventional magnets using
H- ions have been proposed by Martin (Martin,
1987) and a t ITEP in Moscow (Khoroshkov et al.,
1991) and are being considered for hospital use in
northern Italy (Amaldi et al., 1994). The primary
argument for the construction of an H- machine is
the simplicity of beam extraction which is accom-
plished by stripping the electrons off the H- in a very
thin foil target. This leads to a very small beam
divergence, permitting transport of the extracted
beam with narrow gap magnets. But H- machines
have stringent requirements for the vacuum in the
accelerator, due to the possibility of electron strip-
ping by residual gas molecules in the beam pipe.
Magnetic electron stripping is another potential
problem for the H- machine resulting in a maximum

ing a large diameter accelerator ring. The average
beam current could also be rather low for clinical
use.
2.1.3 Linear Accelerators
Linear accelerators are typically characterized by
high energy consumption and a very high beam
intensity, which could produce a potential safety
problem. Nevertheless, a few authors have proposed
this type of accelerator for proton radiotherapy (Boyd
et al., 1982; Hamm et al., 1991). A versatile linear
accelerator was designed a t Los Alamos for pion
therapy and was subsequently calculated to be able
to produce proton beam currents of up to 100 pA a t
an energy of 650 MeV (Boyd et al., 1982). Protons of
approximately 200 MeV could be produced with an
accelerator length of 40 m. Commercial klystrons
from radar equipment could be used to produce
radiofrequency (I$ power for acceleration.
A version of a linac more tailored to the needs of
proton therapy in terms of energy and beam current
was proposed by Hamm et al. (Hamm et al., 1991).
By using side-coupled linac sections for accelerating
protons from 70 to 250 MeV and rf power systems
from medical electron linacs, the facility would be
better adapted for hospital installation and its price
could be reduced.
2.2 Typical Operating Parameters
Depending on the sites to be treated and methods
of beam shaping, energies from 60 MeV to as much
as 250 MeV are necessary, especially if the use of
cross-fired protons for precision treatment in the
head is considered (Larsson and Sarby, 1987). If
ranges in excess of 37 cm are required due to highly
oblique beam entries, or if proton radiography is
planned, energies of 300MeV or more are needed.
Preferably, different beam energies should be avail-
able. If enough beam current is available, this can be
provided by degrading from the maximum energy to
the desired energy with absorbers, although degrad-
ing the primary beam can produce an undesirable
radiation background.
For dynamic beam delivery techniques, which are
discussed in section 2.3, the time structure of the
beam extracted from the accelerator is crucial since
it can determine the accuracy to which the dose can
be controlled. The ideal beam for dynamic therapy is
either a dc beam without time structure, or a beam
with a pulse frequency that is much higher than the
scan frequency. A common way of scanning is with
slow extraction, smearing a spill over several beam
positions. This works well only if the extracted beam
intensity can be well controlled during the spill time.
In order that typical treatments can be completed
in times of approximately one minute, extracted
beam currents in excess of 10 nA are required. For
the largest anticipated field sizes, a current of 10 nA
might result in treatment times of 3 to 4 minutes.
For safety purposes, it is preferred that beam cur-
rents be less than about 100 nA, giving adequate
time to react to beam delivery errors.
Machine stability and reliability are very impor-
tant in the clinical environment. Unscheduled dowil-
time should be comparable to that of x-ray facilities,
typically less than 5%.
2.3 Beam Shaping and Delivery
The merits of protons for radiotherapy lie in the
potential for confinement of the high dose volume to
the target volume and the consequent sparing of
normal tissues outside this volume. It must be the
aim of each beam delivery technique to make opti-
mal use of this potential by providing adequate dose
throughout the target volume while minimizing the
dose to tissues outside the target volume.
Beam delivery techniques are commonly catego-
rized as passive or dynamic. Passive beam delivery is
a method of achieving a spatially uniform dose
distribution by scattering and degrading the pri-
mary proton beam in a set of distributed absorbers to
create the beam diameter, maximum energy and
energy spread needed to deliver uniform dose to the
target a t all depths. An example of such a system is a
rotating propeller with variable thickness blades, as
first discussed by Wilson (Wilson, 1946) in combina-
tion with a separated pair of scatterers (Koehler et
al., 1977). Dynamic beam scanning is a time-
dependent method of achieving a desired dose distri-
bution by magnetically moving the beam across the
target cross-section while dynamically changing the
energy of the beam and, consequently, the depth of
penetration.
Passive beam spreading techniques, because they
irradiate the entire target volume simultaneously,
are both safer and simpler than dynamic techniques.
Dynamic techniques, however, can reduce the dose to
normal tissue and may be preferred in some situa-
tions.
2.3.1 Passive Scattering Techniques
Uniformity of intensity over the useful cross-
section of the beam can be approached by selecting
only the central portion of the gaussian distribution
of a singly scattered beam. This would require a
1arp.e drift distance between the scatterer and the
patient and would result in a low efficiency of beam
usage. Double scattering by a pair of separated
scatterers can increase the fluence in the central
area at the treatment position (Koehler et al., 1977).
with this technique, developed a t Harvard, the first
scattering foil results in a gaussian dose distribution
at the treatment position with a high intensity in the
center of the field. A second composite scatterer,
placed between the first scatterer and the treatment
position, typically has a block on the beam axis
followed by a thin, high-Z scatterer. This combina-
tion reduces the dose in the center and uses the
scattered protons to increase the dose outside the
center. This technique results in a larger homoge-
neous, circular dose distribution a t s specific dis-
tance from the two scatterers and a more efficient
use of the particles, although a large drift distance is
still necessary. The homogeneous dose distribution
must be substantially larger in each direction than
the target volume, and collimators are needed to
shape the proton field to the projected target cross
section. A drift distance of a t least 3 m is recom-
mended between scatterer and patient. This is not a
problem for a fixed, horizontal beam line, but would
result in a large diameter for a gantry beam delivery
system if the scatterers are placed after the last
bending magnet. Smaller drift distances of 1.5 to 2 m
would yield an undesirable increase of surface dose
relative to dose a t depth (Rabin, 1987). The gantry
designed for the Loma Linda facility has a diameter
of 12 m but economizes on the cost of shielding by
arranging most of the magnets in one plane.
A refinement of the double scattering technique
uses a contoured second scatterer of high-Z mate-
rial.3 This main scatterer is combined with a plastic
counterpart, thinner a t the center and with increas-
ing thickness at increasing radii to ensure the same
energy loss of the protons over the entire surface of
the scatterer. If, in addition to the contoured scat-
terer, the scattering and range shifting elements are
placed far upstream of the patient, a sharper dose
fall-off and a higher beam usage efficiency of approxi-
mately 46% result. Beam spreading can also be
achieved by passive magnetic dispersion into a circu-
lar or linear shape (Blosser et al., 1991).
The advantages of passive beam dispersal tech-
niques are safety, simplicity and a lower sensitivity
to the time structure of the proton beam than for any
of the dynamic techniques. However, passive scatter-
ing techniques tend to be sensitive to variations in
beam position. For monitoring and dosimetry, pas-
sive beam spreading results in less stringent condi-
tions on time and spatial resolution. The reduced
flexibility in shaping the dose distribution in three
dimensions is less important for small or regularly
shaped target volumes. For large, irregular target
volumes, the unnecessary exposure of normal tis-
sues adjacent to the target volume can be reduced
with the use of a dynamic multi-leaf collimator in
Personal communication: B. Gottschalk, Massachusetts Gen-
eral Hospital, Boston, MA(1991).
combination with a compensating bolus and a step-
wise reduction of the range of the protons (Chu et al.,
1989).
2.3.2 Wobbling and Beam Scanning
A dynamic technique of beam spreading called
"wobbling" was developed at Berkeley for heavy ion
beams a t the BEVALAC (Chu et al., 1985). By
wobbling, the particles of a beam pulse are smeared
out on rings by the use of a pair of dipole magnets
with fields which vary sinusoidally with time, with a
~ h a s edifference of 90". Several rings of different
iadii and doses are added, depending on the desired
field size, to obtain flat fields of up to 30 cm diameter
with less than 5% dose variation. This technique
economizes on the use of particles, as does double
scattering, but strongly depends on stable pulse
intensities unless large numbers of pulses are used
for each ring. Wobbling uses less material upstream
of the patient, thereby substantially reducing the
adverse effect of beam fragmentation for heavy ions
and the loss of energy inherent in double scattering
methods. This system was in routine use for h e a w
ion radiotherapy a t Berkeley from 1985 to 1992 and
could equally well be used for protons.
Raster scanning, first used a t Uppsala in t h e
1960's ( G r h a n et al., 1985) and later developed a t
Berkeley for heavy ions (Chu et al., 1989), is a
similar but more flexible technique to yield large
homogeneous dose distributions of different shapes.
Instead of scanning on circles, the scans are per-
formed on a rectangular grid with a higher scan
frequency in one direction and a lower frequency in
the direction perpendicular to it. Rectangular fields
of different shapes and sizes can be scanned in this
way giving a field shape more closely related to the
target volume projection, and, therefore, further
reducing the beam particle losses. Up to 40 x 40 cm2
fields could be scanned a t Berkeley with scanning
rates of 40 Hz and 1Hz for the two axes, respectively.
The exact shape of the treatment volume was still
tailored by an individual collimator.
Both wobbling and raster scanning operate with
fixed range modulation. This is achieved by varying
the thickness of absorber material traversed by the
protons. This can be done spatially by using a n
absorber plate with ridge shaped elevations, called a
"ridge filter", or in time, with a rotating absorber
wheel with different thickness sectors. The resulting
dose distribution can be made essentially uniform i n
depth over a distance determined by the thickness of
the volume to be irradiated. Together with collima-
5on of the beam, this results in a single-field dose
distribution which can be conformed to the distal
surface of the target volume contour by the use of a
patient- and beam-specific compensating bolus. Us-
ing fixed range modulation, however, one cannot
avoid exposure of volumes of normal tissue in the
proximal region to approximately the full target
dose. Reduction of dose to the proximal tissues,
resulting in a further improvement of the dose
distribution, can be achieved with the use of a
dynamic multi-leaf collimator, suggested by several
authors but not yet in use (Chu et al., 1989).
2.3.3 Spot Scanning
Spot scanning, often called "voxel" (volume ele-
ment) scanning, represents a further development of
the raster scan described above. Instead of irradiat-
ing rectangular volumes, the scans irradiate only the
defined target volume. Different techniques are fea-
sible for voxel scanning. While varying the speed of
the scanning of a narrow pencil beam is difficult, the
scanning can be done a t constant speed by switching
the beam off and on with a fast "kicker" magnet.
Another solution is to use discrete scanning, which
deposits dose to a voxel and switches the beam off
during the change of parameters for the next voxel
(Kanai et al., 1980). As a three dimensional confor-
mal treatment of a one liter volume with a voxel size
of 5 mm x 5 mm x 5 rnrn needs of the order of 10,000
spots and as the treatment time should be compa-
rable to treatment times for photons, e.g., some
minutes, it requires either fast ramping of the
magnet current or a fast switching magnet, depend-
ing on the time structure of the beam of the accelera-
tor. A voxel or spot scan technique for proton beam
delivery has been developed on the basis of experi-
ence with patient treatments with pions a t PSI
(Pedroni et al., 1989; Blattmann and Coray, 1990;
Blattmann et al., 1990) and a t the Gesellschaft fur
Schwerionenforschung (GSI) in Darmstadt, Ger-
many for light ions (Haberer et al., 1993). The
feasibility of the technique has been demonstrated in
experiments on phantoms. Equivalent techniques
are under development or proposed a t Moscow, by
Ion Beam Applications (IBA) in Belgium and in
Uppsala (Grusell et al., 1991; Jongen, 1991; Khorosh-
kov et al., 1991). For 70 MeV protons, the technique
has been successfully proven in Japan with a pencil
beam of cross-section 1 cm2, cut out of a wide beam
by collimators (Kanai et al., 1983a; Kanai et al.,
1983b).
Even though the techniques cited above appear to
be similar, there are important differences. The
scanning in the direction of the fastest scan is
accomplished with magnetic scanning for all of the
facilities except for IBA. The scan speed along this
fastest direction is 1 m/s or more, making dose
control for each voxel difficult to achieve. Scanning
along the next slower scan axis also uses magnetic
scanning for GSI, Uppsala, Moscow and IBAor range
shifting (depth direction) for the PSI beam. The scan
speed for this axis is a factor of approximately 25
times slower than along the fastest axis. The third
and slowest motion is done either by range shifting
(Uppsala), energy variation (Moscow, also planned
for GSI), rotational movement of the bending mag-
net (IBA) or scanning of the patient (PSI). Common
to all these solutions is the importance of a fast and
reliable monitoring and control system.
Dynamic delivery techniques result in higher flex-
ibility of dose shaping. Depending upon the specific
technique, minimal individual patient hardware such
as collimators and bolus is required. A bolus can be
constructed for specific reasons such as compensat-
ing for structures smaller than the spot size, or a
collimator may be used if the rate of dose fall-off in
the transverse plane should be increased. The ability
to shape the dose distribution to the target volume in
three dimensions results in the potential of reducing
radiation burden to normal tissues and hence increas-
ing the therapeutic gain (Urie and Goitein, 1989) or
reducing the number of beam ports and conse-
quently reducing treatment complexity. Dynamic
beam delivery techniques permit the design of a
gantry with a much smaller diameter, particularly in
the case where the patient couch is mounted eccentri-
cally (Pedroni et al., 1989). Parallel scanning in a
cartesian grid makes treatment with abutting fields
somewhat easier. The sharp dose fall-off of proton
beams is a potential source of cold or hot regions near
the junction of abutting fields. The sharpness of the
dose fall-off can be modified in these regions to
reduce the sensitivity to positioning errors or patient
movement.
A topic which needs special attention if dynamic
beam delivery techniques are used, is the effect of
patient or organ motion on the dose distribution.
Although minimizing patient and target volume
movement during treatment is always important for
proton radiotherapy, when using passive beam
spreading techniques, such movements will result in
a geometric miss only if the margins allowed around
the target volume are too small to account for this
motion. On the other hand, when using dynamic
beam delivery, only a portion of the target is being
irradiated a t any moment, so movements of the
patient or target volume during treatment can lead
to unacceptable dose inhomogeneities inside the
target volume (Levin et al., 1988; Phillips et al..
1992).
 3. Pertinent Quantities and Units
3.1. Physical Quantities
The following is a listing of physical quantities
relating to proton therapy. Complete definitions can
be found in ICRU Report 33 (ICRU, 1980)which also
contains definitions of other radiation quantities
that relate to neutral particles.
Because of the discrete nature 3f the interaction
of ionizing radiation with matter, the phenomena
described by radiation quantities are subject to
fluctuations. In many practical situations, the devia-
tions from the mean values that are represented by
non-stochastic quantities are negligible and unless
otherwise indicated, the quantities listed here are
non-stochastic. In some instances, statistical devia-
tions are important and two stochastic quantities
that are subject to probability distributions are
defined.
Values of radiometric and dosimetric quantities
are generally a function of time and their differentia-
tion with respect to time results in what are usually
termed rate quantities. Thus the absorbed dose rate
is given by
and conversely the absorbed dose delivered in a time
interval T is
Because of this simple relation, it is sufficient to
define either the differential or the integral type of
quantity.
Many of the definitions incorporate products of
quantities. For a monoenergetic beam, the energy
flux, R, can be defined by
where N is the particle flux and E the energy of the
particles. Frequently, the beam of particles has a
distribution of energies in which case the energy flux
becomes
The units employed are those of the International
System (SI) with the special name gray (Gy) for 1J
kg-1 reserved for dosimetric quantities. Certain units
(electron volt, minute, etc.) may be employed with
the SI and it is also customary to employ multiples
and submultiples of units (e.g. keV mm-l). The units
of the quantities defined below are indicated in
parentheses.
3.2 Radiometric Quantities
The (particle) flux is
where cW may refer to the number of particles
emitted, transmitted or absorbed (e.g., in a beam of
protons) in time dt.
The energy flux, as indicated above, is defined as
for a monoenergetic beam, where E is the particle
energy (e.g., R is the energy transmitted in a proton
beam per unit of time).
The other radiometric quantities, as well as the
dosimetric quantities, are point functions and, in
general, their measurement requires instruments in
which the sensitive elements have small geometric
dimensions.
The (particle) fluence is
where cW is the number of particles traversing a
sphere of cross-sectional area da centered a t t h e
point of interest. In the case of a unidirectional
beam, da is simply an area perpendicular to t h e
beam.
The (particle) fluence rate
has been symbolized by the lower case Greek letter
rather than by & .
Analogously the energy fluence is
and the energy fluence rate is
3.3 Interaction Coefficients
The mass electronic stopping power is
Slp = (llP)(dEldx),(Jm2kg-' or MeV m2kg-') (3.11)
where dE is the mean energy lost by a charged
particle in electronic collisions while traversing the
distance dx in a material of density p. In this
notation, mass electronic stopping includes all inter-
actions with atomic electrons either individually,
denoted eiectronic in the notation of ICRU Report 49

notation of ICRU Report 49. Radiative transitions
and nuclear interactions are not included.
pdx is termed the areal density, S the linear
stopping power.
Slp depends on the composition of the material and
on the nature and energy of the charged particle. In
the range of energies that are employed in therapy,
protons expend most of their energy in electronic
collisions with a negligible loss by radiative pro-
cesses. These radiative processes bemine significant
only at energies exceeding 1 GeV. The energy in-
volved in elastic and non-elastic nuclear interactions
is not included in S 1p.
The linear energy transfer or restricted linear
collision stopping power is defined by:
where dE is the energy expended by a particle
traversing a distance dx reduced by the kinetic
energy of electrons with energy greater than A. This
quantity is usually termed the LET. L, (also written
as L ) is equal to S.
The mean energy expended in a gas per ion pair
formed is
W = EIN, (J,eV) (3.13)
where E is the total energy of a charged particle and
N is the mean number of ion pairs it produces when
it is completely stopped in a gas. W depends on the
composition of the gas and on particle type and
energy. The dependence on energy may require the
differential quantity w(E) = d E l d N in ionization
measurements.
3.4 Dosimetric Quantities
A very useful quantity called cema (converted
energy per unit mass) has been introduced (Kellerer
et al., 1992) and is defined as
C = @SIP. (J kg-' or Gy) (3.14)
Cema, C , represents the energy lost by heavy
charged particles through electronic interactions in
a manner analogous to the way in which herma
represents the energy transferred to directly ioniz-
ing particles by indirectly ionizing particles (Roesch,
1958; ICRU, 1980). This quantity, which is equal to
the energy lost in electronic interactions by a fluence
of charged particles per unit mass of an irradiated
specified material, may refer to a point inside an-
other material or to a material in free space.
The absorbed dose is defined as
D = dZldn7, (J kg-' or Gy) (3.15)
where d is the mean energy imparted to matter of
mass dm. The bar over E is employed to provide a
distinction from the stochastic quantities listed be-
low.
Due to energy transport by secondary electrons
(delta rays), absorbed dose cannot be identical to
cema. Although the maximum delta ray range for
250 MeV protons in tissue is approximately 2.5 mm:
the vast majority of delta rays produced in protor!
therapy beams have ranges considerably smaller
than 1 mm. Thus, complete delta ray equilibrium
can be assumed to exist throughout irradiated mat-
ter and absorbed dose and cema can be considered t~
be numerically equal.
3.5 Stochastic Quantities
Statistical fluctuations of the energy imparted by
individual protons are large in volumes comparable
to those of cells, their nuclei or smaller biological
units. These fluctuations are of negligible impor-
tance in clinical absorbed dose measurements but
will be considered in more detail in Part I1 of this
Report. The stochastic quantities refer to observed
rather than expected (mean) values of energy.
The lineal energy is defined by
where E is the energy imparted in a n energy absorp-
tion event to the matter in a volume of interest, and 2
is the mean chord length in the volume. The volume
is usually taken to be a sphere of diameter d, and
then 1 = (213) d. The term (energy absorption) event
refers to energy deposition by one charged particle or
a group of correlated particles (e.g., a proton and its
delta rays).
A probability density distribution fly), describes
the distribution ofy. The mean value of this distribu-
tion is referred to as the frequency average and is
given by:
The dose weighted average ofy is written as:
The specific energy is
where E is the energy imparted to matter of mass m.
z may be due to one or more energy deposition
events. The probability (density) distribution ffz)
depends on m.
The mean value of z is
where is the mean absorbed dose in the mass for
which f- (z) is determined. It should be noted that
lim,,,,oz = D.
 4. Proton Interactions with Matter
4.1 ~ l e c t r o m a ~ n e taincd Nuclear
Interactions
Protons lose their energy to matter primarily
through electromagnetic interactions with atomic
Protons have a mass which is large com-
pared to the mass of the electrons, hence they lose
only a small fraction of their energy in a single
interaction (at most 4 m / M = 0.0022, where m is the
electron mass and M is the proton mass) and they
are deflected by only small angles in each interac-
tion. In general, the proton interactions with matter
can be divided into three categories (Bichsel, 1968):
interactions with the individual electrons of atoms,
interactions with the nucleus and interactions with
the atoms as a whole. The latter occur only a t very
low energies and will not be considered in this
Report. Nuclear interactions include inelastic scatter-
ing, Rutherford scattering and nuclear reactions.
Although this document will consider the effects of
nuclear interactions on dosimetry, it will concentrate
mostly on the description of proton interactions with
electrons of atoms and molecules, since a t therapy
energies, these processes dominate.
The primary dosimetric quantity of interest for
radiotherapy is the absorbed dose, which is defined
as the mean energy imparted per unit mass of
material (see Equation 3.15). For charged particles,
the fundamental energy loss quantity is cema (Equa-
tion 3.14) although, as discussed in Section 3.4
above, it is reasonable to assume that absorbed dose is
numerically equal to cema. The processes by which
protons slow down and deposit energy along their tracks
determine the distribution of absorbed dose in the
patient.
In the slowing down of protons in any material,
some of the molecules of the material are excited, but
the more important effect is ionization. If the kinetic
energy imparted to an electron in the ionization
process is sufficient for it, in turn, to cause subse-
quent ionization, it is called a 6 ray. The average
energy required to produce an ion pair, referred to as
W, is a n important dosimetric quantity, particularly
for ionization chambers. Section 5 of this Report will
deal with Win more detail.
Application of the ionization phenomenon can be
seen in the use of semiconductors or gases as particle
detectors. When charged particles lose energy in a
semiconductor, electrons are excited from the va-
lence band to the conduction band, thereby changing
the conduction properties of the material. With the
application of an applied potential, current flow cpr,
be used as a measure of the energy deposited by
charged particles. The operation of ionization cham-
bers also depends on current measurement. Ioniza-
tions and excitations in photographic emulsions can
form a permanent record of the passage of charged
particles following the development process. In some
materials (scintillators), ionizations and excitations
can produce light quanta in the visible range.
4.2 Energy Deposition
4.2.1 Average Energy Loss
The proton mass stopping power of a medium,
(Slp), is defined in Equation 3.11. Frequently, the
unit MeV cm2g-l is used, where the areal density (in
g ~ m - is~ )defined as the product of density p and
absorber thickness t. In the continuous-slowing-down
approximation (csda), the range of a particle is given
by R = $[S(E)I-ldE. Due to the fact that each particle
experiences a different set of interactions, a group of
particles of the same initial energy has a distribution
of energies after traversing a thickness of absorber
(energy straggling) and a resulting distribution of
depths a t which the particles stop (range straggling).
Radiation dosimetry and radiobiological modeling
are concerned with the deposition of energy along
the track of single charged particles. This is specified
by the restricted stopping power or LET (linear
energy transfer), (Equation 3.12), defined as the
mean energy loss per unit pathlength due to colli-
sions involving energy transfers that are smaller
than some cutoff. This cutoff is chosen to correspond
to the energy of secondary electrons with ranges
which are equal to a relevant radius around the
particle track. This restricted stopping power speci-
fies an average energy loss and is a non-stochastic
quantity. Actual energy deposition is specified by the
stochastic quantities of microdosimetry.
Proton energy loss mechanisms may be considered
in three separate proton energy regions: Low energy,
below 10-5 Me2 (where Me2 refers to the particle's
rest mass), intermediate energy, and high energy,
above Me2. In each of these energy regions, qualita-
tively different energy loss phenomena dominate.
At low energies, which for protons is below about
0.01 MeV, elastic collisions between the particles and
whole atoms of the medium are important. Since this
region corresponds to less than one pm of the proton
trajectory in tissue, it can be ignored for purposes of
proton dosimetry for radiotherapy.
At increasing energy, the nuclear interactions
become more important. In the intermediate energy
range, the probability of nuclear events is small
compared to the probability of electron interactions,
although each nuclear reaction can transfer a signifi-
cant portion of the proton energy to the medium. For
very high energy protons of 1,000 MeV and above,
the probability of suffering a nuclear interaction
before the end of the range (as defined by electron
interactions) is high, hence the entire concept of
range becomes less meaningful. In addition, proton
bremsstrahlung begins to contribute to the energy
loss and bulk properties of the medium (dielectric
constant, etc.) begin to influence the collision process
and modify the stopping power via the phenomenon
called the density or polarization effect (Sternhei-
mer, 1966).
In the energy range between 0.01 MeV and 250
MeV, where the residual range of the protons in
tissue is between a micron and approximately 35 cm,
interactions with electrons are dominant. Even a t
these energies, however, one cannot ignore the ef-
fects of nuclear interactions. For tissue-equivalent
material, the probability that protons will undergo a
nuclear interaction while traversing a path length
segment of 1 g is of the order of 1% (ICRU,
1993). At a depth of 20 cm, approximately 1in 4 of
the protons will have suffered a nuclear interaction.
This will contribute a background of nuclear interac-
tion products (Kliauga et al., 1978)which can modify
the biological effect of the proton radiation (Hall et
al., 1978; Verhey et al., 1979). Nuclear interactions
affect the identity and energy distribution of the
secondary particles and decrease the number of
primary protons in the beam.
The mass electronic stopping power of matter for
protons, calculated purely on the basis of electronic
interactions (ICRU, 1993), is given by:
where ze and p refer to the charge and the speed (in
units of c , the velocity of light) of the incident proton,
m is the mass of the electron, 2,N and I refer to the
mean nuclear charge, the mean density and the
mean excitation energy of the atoms of the medium
and C j / Z represents the shell corrections which are
most important a t low velocities (Bichsel, 1968;
Bichsel, 1992). For protons of very high energy (with
p > 0.9), an additional term needs to be added to the
stopping power equation which corrects for the den-
sity effect (Sternheimer, 1966). ICRU Report 49
(ICRU, 1993) tabulates proton stopping powers in
the entire energy range of interest for radiotherapy.
4.2.2 Microdosimetric Concepts
Although the average energy loss per unit mass,
the absorbed dose, is a useful macroscopic concept
defining the ambient energy concentration a t the
site of interest, the microscopic local energy density
or specific energy z = d m as defined in Equation
3.19, may reach far higher values for very small
volumes. The specific energy in microscopic domains
is responsible for biological and chemical radiation
effects which are generally produced more efficiently
by high-LET rather than low-LET radiations. At the
energies employed in therapy, protons can be re-
garded as low-LET radiation. However, a small
portion of the energy loss of protons is through
nuclear interactions, leading to a small'probability of
large local depositions of energy. Measurements of
the dose-weighted lineal energy distribution (see
Equation 3.18) in a clinical proton beam indicate
that not more than about 2% of the absorbed dose is
due to these types of events (Kliauga et al., 1978) for
energies less than 250 MeV but, because of their
high biological effectiveness, their contribution is not
negligible. This subject will be considered in Part I1
of this Report.
4.3 Factors which Affect Proton Beam
Characteristics
The interactions of protons with matter can be
described statistically since they are made up of a
very large number of events, each of which is respon-
sible for only a small amount of energy loss (average
energy losses per collision are of the order of 100 eV).
Nuclear events result in the loss of the primary
protons and the production of secondary particles.
For a monoenergetic parallel beam of protons strik-
ing matter, one will observe (Bichsel and Hiraoka,
1989):
1. the production of secondary electrons (6 rays)
from ionizations,
2. the loss of primary protons before the end of
range due to nuclear interactions and the result-
ing creation of secondary particles, a s well as
the following phenomena, all of which depend
on both the thickness of material traversed and
the incident energy, namely,
3. range straggling, i.e., a distribution of stopping
depths around a mean which can be calculated
and which has a width that depends on the
material of the absorber,
4. energy straggling, i.e., a distribution of ener-
gies around a mean value which can be deter-
mined in an energy loss calculation, and
5. a n angular distribution around the central axis
with a width which can be calculated from
multiple scattering theory (Bethe, 1953; Bich-
sel et al., 1982).
Although proton stopping power can be calculated
a t any given energy, the presence of nuclear events
and the difficulty of measuring the beam characteris-
tics lead one to conclude that absorbed dose in the
patient and clinical proton beam parameters should
be based on measurements made in materials as
similar as possible to the patient. Section 7 of this
Report recommends dosimetry procedures which
should be used to calculate absorbed dose in the
pa t.ient.
 5. Determination of Proton Absorbed Dose in Reference Conditions

5.1 General Considerations for

Proton Dosimetry
The principal transfer of energy from protons to a
involves a two stage process: (1) energy loss
by the protons in ionizations and excitations and (2)
the subsequent absorption of this energy which is
largely transmitted by 6 radiation. In the case of
uncharged particles, one distinguishes between
kerma (the energy transferred to charged particles)
and the absorbed dose. In the case of charged par-
ticles, the distinction is between cema and absorbed
dose.
For protons of energy E and fluence @, the cema is
0.0
given by 0 5 10 15 20 25 30 35 40 45
W a t e r Depth / cm
Fig. 5.1. Relative absorbed dose values due to 250 MeV
and for a fluence distributed in energy, protons incident on a water slab. Total absorbed dose values and
those due only to protons are shown. The proton-only dose values
are multiplied by 0.9 for ease of viewing. The vertical line
indicates the depth of csda range for primary protons.
where S(E),lp is the mass electronic stopping power
for medium m in units of J m2kg-l, @ is the fluence
in m-2 and @dE)is the fluence in m-2 J-l. Proton neutrons is demonstrated in Figure 5.3. The values
generated secondary electrons are of low energy and shown in Figures 5.1 through 5.3 were calculated
short range, so cema and absorbed dose are approxi- using the LAHET code (Prael and Liechtenstein,
mately equal, similar to the situation of kerma and 1989).For these calculations, a non-divergent beam
absorbed dose for indirectly ionizing radiation when of protons was allowed to impinge upon an infinite
charged particle equilibrium exists. The greater the slab of water. The resulting dose is equivalent to that
validity of this approximation, the simpler the inter- which would be obtained in a small region in the
pretation of measurements and the more accurate center of the water slab due to bombardment with a
the prediction of dosimetric quantities. In the follow- uniform parallel fluence of protons. Since nuclear
ing, cema and absorbed dose will generally be as- interactions are modeled by LAHET using an intra-
sumed to be equal.
Below a few hundred MeV proton energy, elastic
and nonelastic nuclear interactions occur but, with a
few exceptions mentioned below, they are of second-
ary importance for determining the absorbed dose in
clinical situations. Figure 5.1 shows the relative 0 Protons
absorbed dose as a function of depth for a 250 MeV
plane parallel beam of protons impinging upon a Alpha
x Nuclear Recmls
water slab ("planar fluence").In Figure 5.2, absorbed
dose values normalized to the total absorbed dose at
that depth are plotted as a function of depth in water
for primary and secondary protons, mass two and
mass three particles, alpha particles and heavier
nuclear recoils. Nonelectronic interactions contrib-
ute less than a few percent to the absorbed dose
except near and just beyond the primary proton
range. Amongst the nuclear processes that occur,
neutron production has a significant impact on ab- Fig. 5.2. Absorbed dose values due to 250 MeV protons
sorbed dose due to the large mean free path of these incident on a water slab as a function of depth. The values shown
are for protons (primary and secondary), mass two and three
neutrons combined with the production of heavy particles, alpha particles and nuclear recoils, and are normalized
charged particles generated by subsequent neutron to the total absorbed dose a t that depth. The vertical line indicates
interactions. The significant fluence of secondary the depth of csda range for the primary protons.
Y
E = 250 MeV, P l a n a r Fluence
I I P 1 I I
to0-
10-I- Proions
Depth = 2 mn
loJ -
lo-*-
lod I I I I I I
0 50 100 150 200
Proton Energy / MeV
I
Ep= 250 MeV, P l a n a r Fluence
I I
I I
Neutron8
Depth = 2 mn
lo6 I 1 I I I I
0 50 100 150 200
N e u t r o n E n e r g y / MeV
Fig. 5.3. Relative proton and neutron spectral fluence per
incident proton a t a depth of 2 cm for a parallel beam of 250-MeV
protons incident on a water slab.
nuclear cascade, which is not entirely appropriate in
this energy region, these results only estimate pos-
sible energy depositions from nuclear interactions.
However, the results do suggest that the cema
approximation will generally be applicable for inter-
preting the response of most instruments and dosim-
eters in the region containing primary protons.
Nuclear secondaries may become important when
consideration of the biological effectiveness elevates
their dosimetric impact. As indicated in Figure 5.2,
this occurs near the end of the proton range. In
addition to the potential biological effect due to
nuclear secondaries for a range modulated proton
beam, slowing protons near the Bragg peak may be
expected to produce an enhanced biological effect.
Absorbed dose determination for energetic proton
beams is an inherently simpler process than for
indirectly ionizing radiation or for electrons. The
dosimetry of indirectly ionizing radiations is compli-
cated by the sensitivity of the response to the materi-
als that compose the dosimeter. Therefore, the el-
emental composition of dosimeters used for such
radiations should closely approximate that of the
material in which absorbed dose is to be determined,
such as water. Although electrons are directly ioniz-
ing, their dosimetry is also difficult since they un-
dergo frequent scattering due to electron-electron
collisions and lose considerable energy from single
I
I
event radiative processes. Therefore, the dosimeter
for use in electron beams must respond to a large
range of electron energies as well as a spectrum of
photon energies. Not only must the dosimeter repli-
cate the elemental composition of the material in
which absorbed dose is to be determined, but the
dosimeter must not perturb the fluence of electrons
and photons traversing the dosimeter. For both
electrons and photons, the response of the dosimeter
I is dependent on the surrounding media.
250 300 When proton absorbed dose is determined, it is
usually measured in some material that differs from
I
the material of interest. For example, measurements
might be made in a water medium using a gas-filled
ionization chamber constructed with walls of some
other material such as plastic. The absorbed dose to
soft tissue is then inferred from the response of the
dosimeter which is not composed of tissue. I n this
case, the ionization produced in the filling gas must
be related to the absorbed dose in the material of
interest. To the degree that the cema approximation
applies, interactions in the chamber walls and sur-
rounding media produce no direct response in the
250 3
4 gas. The relationship between absorbed dose in the
gas and absorbed dose in the medium of interest is
determined by the ratio of the mass electronic stop
ping powers for the two materials. If the ratio of
these stopping powers is independent of proton
energy, the ionization chamber response determines
the absorbed dose to water or tissue in a simple and
straightforward manner.
The response of a radiation detector to a proton
beam is also influenced by the geometry of the
detector. This is due to the fact that the e n e r z
deposited by a particle in the detector depends on
both the effective stopping power of the medium and
the path length of the particle in the medium. As an
example, the response of cylindrical and parallel
plate ionization chambers would be expected to differ
due to dserences in mean path length (Bichsel,
1996).
5.2 Fluence Measurements with a
Faraday Cup
A Faraday cup is a device which can be used to
determine the number of protons in a beam. Protons
that reach the thick absorber inside the Faraday cup
produce a net charge proportional to the number of
protons, Q = N e, where Q is the charge collected.
and e is the charge per proton. The electrically
insulated and conducting proton beam absorber must
be thick enough to stop all primary protons and
proton-produced secondary charged particles in the
absorber. Such instruments are commonly used to
meascre the current in charged-particle beams at
accelerators. Proton currents can be determined
accurately with this device.
A potential error in this measurement can occur
from particle fluence not due to the proton beam.
Interactions of protons with upstream absorbers
produce charged spallation products and electrons
that may add to, or subtract from, the measured
charge. Charged products generated in the sensing
absorber, usually electrons, may escape the absorber
and modify the response. Appropriate use of thin
entrance foils, vacuum environs, and trapping elec-
tromagnetic fields surrounding the sensing absorber
minimizes effects due to secondary electrons (Verhey
et al., 1979; Vynckier et al., 1984; Kacperek and
Bonnett, 1989).
A more subtle problem concerns energetic protons
and other heavier charged particles produced near
the periphery of the absorber by fast neutrons gener-
ated by proton interactions in the upstream portion
of the sensing absorber. The use of non-hydrogenous
high-Z absorbers with concomitantly small particle
production cross sections minimizes this effect.
Finally, if the bombarding proton beam is com-
pletely stopped in the Faraday cup and is uniformly
distributed in a known area a, the measurement
yields a proton fluence or fluence rate. The cema in
medium m for this beam condition is given by:
(5.3)
where (310)=
J @p, (Ep)(S(EpYp)dE
where QPE is the proton fluence and N is the number
of protons in beam area a.
Under the cema approximation, the Faraday cup
can be used to calibrate a dosimeter in terms of
Vacuum chamber
Ceramic insulator Beam absorbing cup
Fig. 5.4. Schematic view of a Faraday cup suitable for energetic proton beam fluence daleminations (Raju et al., 1969).
absorbed dose to material m. For dosimeter calibra-
tion, care must be exercised to ensure that the proton
spectral fluence bombarding the Faraday cup is well
known. In addition, the uniformity of the proton
fluence in area a must be confirmed as the Faraday
cup responds indiscriminately to all charged par-
ticles entering the collection mass.
Figure 5.4 shows a schematic view of a typical
device employed for such measurements (Raju et al..
1969).
5.3 Absorbed Dose Measurements
w i t h a Calorimeter
Unlike measurements based upon the products
produced by the interaction of ionizing radiation
with matter, e g . , ionizations, a calorimetric measure-
ment is a direct determination of the energy im-
parted to a sensing element as indicated by a tem-
perature change. Assuming that all the deposited
energy is thermalized, absorbed dose may thus be
directly determined in a calorimetric measurement.
Knowledge of such radiation parameters as the
mean energy needed to create an ion pair, W, or the
chemical yield per unit energy deposited, G, and of
the dependence of these quantities on ionizing par-
ticle species and energy, is not required. A calorimet-
ric absorbed dose determination can provide an
independent confirmation of ionization determina-
tions that is especially important for absolute re-
sults. A complete discussion of calorimetry measure-
ments may be found in Domen (Domen, 19861, while
Attix (Attix, 1986) provides an excellent descriptive
introduction.
Implementation of calorimetric determinations in-
volves the measurement of the temperature change
in a mass of material resulting from energy imparted
by ionizing radiation. Many such devices operate in
To electrometer
t
Vacuum window
an adiabatic manner. A core of' material is sur-
rounded by one or more insulating jackets. Before,
during, and after irradiation, the core and surround-
ing jackets, which are irradiated simultaneously, are
maintained a t equal temperature. The net core
temperature change, relative to the pre-irradiation
period, is then proportional to the energy deposited
by the ionizing radiation. To ensure that the signal
derives from the material of interest, even for large
imparted energy density, a core mass of several
grams is usually employed. The imparted energy per
unit mass is then representative of the average
absorbed dose to the core material. Conversion of the
measured temperature change to energy imparted is
accomplished by knowledge of the mass of the core,
combined with either calibration of the temperature
response of the core with resistive heating or knowl-
edge of the core specific heat value. Adiabatic calorim-
eters usually use homogeneous core and jacket mate-
rials such as graphite (Domen and Lamperti, 1974)
or A-150 tissue equivalent plastic (Mcdonald et al.,
1976). The temperature change is typically of the
order of O C per Gy. Thermistors are metal oxide
semi-conductors with a negative temperature coeffi-
cient which can be used in conjunction with a
Wheatstone Bridge to conveniently determine this
temperature change with great precision.
Water calorimeters have been developed to pro-
vide a more direct determination of absorbed dose to
water (Domen, 1980). These devices are operated in
a non-adiabatic manner and the temperature in a
small region of the water calorimeter surrounding
the thermistor temperature-sensing element is di-
rectly measured. If conduction and convection heat
losses are minimal, the temperature rise is directly
related to the energy imparted per unit mass near
the measurement point. Knowledge of the thermal
heat capacity is assumed in this determination.
Schulz and co-workers have developed a water-based
portable calorimeter for absorbed dose determina-
tions in photon, electron and, more recently, in
proton beams (Schulz et al., 1987; Schulz et al., 1991;
Schulz et al., 1992). Domen has provided a thorough
discussion of the implementation of water-based
calorimetry (Dornen, 1994).
Although calorimetry avoids difficulties associated
with determination or knowledge of the radiation-
specific parameters, technical problems may limit
the accuracy of the measurement. Perhaps the most
significant is the conversion of deposited energy into
non-thermal processes such as chemical reactions.
These reactions may create or absorb heat, ie., they
can be either exothermic or endothermic. In the
latter case, one can define a thermal defect as the
fraction of the energy imparted that does not appear
as heat in the calorimeter material. For example, the
thermal defect for A-150 plastic is reported to be
about 4% 2 1.5% (Fleming and Glass, 1969; Mcdon-
ald and Goodman, 1982; Schulz et al., 1990). Since
A-150 plastic is a mixture of several materials, the
thermal defect may be sensitive to the manufadur-
ing and curing process as well as the radiation
history of the sample. Not surprisingly, the thermal
defect in water can be sensitive to absorbed impuri-
ties. Thermal defect values of several percent are
possible. Investigations by Domen (1994) and Schulz
et al. (1992) indicate that by using nitrogen-purged
high resistivity water, the thermal defect can be
made small.
Beyond the difficulty of the thermal defect, a
number of other technical problems are frequently
encountered. For water calorimeters operated near
room temperature, convection currents are a recur-
ring problem. To achieve an adequate signal-to-noise
ratio for the small induced changes in thermistor
resistance from radiative heating of the water bath,
low power dissipation in the thermistor is essential.
For adiabatic calorimeters, the conversion from ab-
sorbed dose in the core to absorbed dose in water is
accomplished by using the mass electronic stopping
power ratio between the core material and water.
Hence, knowledge of the energy spectrum of the
directly ionizing particles creating the heat is re-
quired. Since the calorimeter should be operated
adiabatically, careful control of heat exchange be-
tween the core and jacket is needed. For measure-
ments near the Bragg peak, correction should be
made to the stopping power due to energy loss of the
particles in the core.
5.4 Absorbed Dose Measurements with an
Ionization Chamber
5.4.1 Introduction
The field of radiation measurements is dominated
by gas ionization techniques. Gas filled ionization
detectors are widely used in proton radiotherapy.
Applications include: monitoring proton beam inten-
sity, determination of spatial beam location, verifica-
tion of the uniformity of proton intensity following
lateral dispersion, absorbed dose determinations in
phantoms, etc. For each application, ionization detec-
tors bring a complement of advantages and disadvan-
tages.
5.4.2 Conceptual Description of Absorbed
Dose Determination
As already discussed, protons are, for the most
part, directly ionizing, losing energy primarily
through electronic interactions. This transfer of en-
ergy to secondary charged particles is defined by
cema. Subsequent interactions by these secondary
charged particles impart energy to the media, de-
scribed by absorbed dose. Even for very energetic
protons, secondary electrons impart their energy to
the media close to their point of creation so cema and
absorbed dose are effectively equal. Therefore, the
absorbed dose, D ,and cema, C , in the gas of the
ionization chamber cavity are given by,
Dg,i(E) C , i ( E )= @JE>(S(E)Ip),j, (5.4)
where i indicates particle type, E is the particle
energy, and (S(E)Ip),,i is the mass electronic stopping
power of gas g for particles of type i.
The measured quantity is the charge produced in
the gas by directly ionizing particles and it is related
to the absorbed dose in the gas by
where Jg,&E)is the charge produced per unit mass of
gas and w,{E) is the differential mean energy re-
quired by particles of type i with energy E to produce
an ion pair in the gas. The use of differential w or
integral W values is discussed in Section 5.4.3.2.
Including all particle species and energies, the charge
produced per unit mass of gas is given by
Jgcan be related to D,by
where E represents a value averaged over particle
type and energy. Absorbed dose in the gas and
absorbed dose in another medium are related by the
ratio of mass electronic stopping powers and thus the
absorbed dose in the medium, Dm,is given by
where m indicates the medium of interest.
5.4.3 Interpretation of Ionization
Chamber Response
Inherent in these considerations is the assumption
that the gas ionization is produced only by the
primary particles, that is, that particles generated in
the surrounding medium, e.g., the chamber wall
material, contribute no additional energy deposi-
tions to the gas. Given the above assumption, two
additional problems relative to the determination of
absorbed dose must still be addressed: (1)the impli-
cations of the dependence of the stopping power on
medium and particle type and ( 2 ) the effect of
replacement of the particle- and energy-dependent
w(E) value by a mean value. These are discussed in
the sections below.
5.4.3.1 Considerations of Stopping Power.
The dependence of the stopping power on medium
and particle type is the easier problem to address.
Recognizing that the primary particles are protons
and that for energies below a few hundred MeV,
proton energy losses are due primarily to electronic
interactions, contributions to ionization by charged
particles other than protons can be ignored. Hence
the absorbed dose to the medium is related to
ionization in the gas by
Noting that the dependence of this ratio of integrals
upon particle energy and material atomic number
becomes significant only a t particle speeds approach-
ing that of the orbital electrons (Evans, 1955), a
mean value may be used to approximate the spectral
integration,
Therefore, the absorbed dose to the medium is given
by
where s,, represents an effective value that takes
into account the proton energy spectrum. The method
for determining the effective value is discussed in
Section 7.5.
Use of the effective stopping power approximation
is unnecessary if the ionization chamber filling gas
and detector material are identical in elemental
composition and furthermore identical to the mate-
rial under study, e.g., water. In that case, there is no
stopping power correction and any energy deposi-
tions in the gas not due to primary particles interact-
ing with the gas or due to particles emanating from
the surrounding medium are properly incorporated
into the gas response, i.e., the detector functions a s a
homogeneous Bragg-Gray cavity.
In general, no material exactly matching tissue or
water is available for ionization chamber construc-
tion. Note that a water calorimeter very closely
approximates a perfect soft tissue dosimeter for
protons. A solid tissue substitute is A-150 tissue
ey~ivalentplastic which is commonly employed for
chamber construction. This material closely matches
the hydrogen content of muscle but replaces most
oxygen with carbon (Smathers et al., 1977). so-called
methaso- or propane-based tissue equivalent gases
 5.1 -Elemental composition of several solids, liquids und gases. The clensity ofgases is for a pressure of I u t m
TABLE
and a temperature of 20iC
Percent elemental mass Denxty
Substance H C NO 0 Other (e/cm31 Reference

Standards
standard man 1.5 Ca, 1.0 P; 0.8 S + K + Ca. ICRP (1995)
muscle (striated). l.lNa+Mg+P+S+Na+ ICRU (1964)
Mg+K+Ca+Cl
muscle (skeletal) O.lNa,O.2P,0.3S,O.lC1,0.4K ICRU (1989)
bone (compact) 7.0 P, 14.7 Ca ICRU (1964)
A-150 plastic 1.8 C1,1.7 F Smathers et al. (1977)
carbon-graphite
LiF 26.8 Li, 73.2 F ICRU (1984b)
Lucite (C PH8 0 2 ) n ICRU (198413)
Nylon ICRU (1977)
Polyethylene ICRU (1984b)
Polystyrene ICRU (1984b)
Teflon ICRU (1984b)
Liquids
Water
Muscle equivalent Goodman (1969)
Gases
Air ICRU (1984b)
Muscle equivalent-methane Rossi and Failla (1956)
Muscle e&ivalent-propane 10.3 56.9 3.5 29.3 0.00183 Srdoc (1970)

have compositions intermediate to tissue and A-150 cially near the Bragg peak. Carbon and air are more
plastic. For comparison purposes, Table 5.1 lists the closely matched. The opposite situation occurs for
elemental compositions of several important dosimet- TE-methane filling as the concentration of hydrogen
ric materials (Rossi and Failla, 1956; ICRP, 1959; is well matched to that of water, A-150, and muscle.
ICRU, 1964; Goodman, 1969; Srdoc, 1970; ICRU, Here the ratio of proton stopping powers for carbon
1977; Smathers et al., 1977; ICRU, 198410; ICRU, to TE-methane varies by about 6% over the energy
1989). As can be observed, the elemental composi- range of 1MeV to 250 MeV and by about 1%over the
tions of common dosimetric materials only approxi- range of 10 MeV to 250 MeV. Conversely, the A-150 to
mate the values for soft tissue. The potential error
introduced by using these materials and the afore-
mentioned approximation to the conversion from
ionization to absorbed dose is directly related to the
relative mass electronic stopping power values.
The validity of using a mean value for the stopping
power ratio for the material and filling gas is demon-
strated by considering the variation of this ratio for
various materials, proton energies and filling gases.
Such values are plotted in Figures 5.5 and 5.6 for the
materials muscle, water, A-150 tissue equivalent
plastic, and carbon and for a filling of air or methane- --- Water
based tissue equivalent gas mixture (Rossi and
Muscle
Failla, 1956) versus proton energy. Stopping power
data used to calculate these ratios were taken from ---. Carbon
the ICRU tabulation (ICRU, 1993) and are summa-
rized in Table 5.2. As seen in the figures, the varia-
tions in the ratios of stopping powers with energy are 1.00
primarily a t energies below 10 MeV.
The ratio of stopping powers between 1 and 250
MeV proton energy varies by less than 6% and by
about 1.1% for A-150 plastic to air and carbon to air,
respectively. In the range of 10 MeV to 250 MeV, the
same ratios vary by 2% and 0.9%, respectively. The Proton E n e r g y / MeV
presence of hydrogen in A-150, water, and muscle Fig. 5.5. Ratios of the mass electronic stopping power of
increases the stopping power relative to air, espe- several materids to air plotted versus proton energy.
 ,- - - - - - - - - - - - - - - - - - - - - - -
I _ / _ _ _ _ -
-
$ ----- A-I50
2 -
?
WJ 0.92- --- Water
b energies less than E.
Muscte
P r o t o n Energy / MeV
Fig. 5.6. Ratios of the mass electronic stopping power of
several materials to TE-methane plotted versus proton energy.
TE-methane ratio varies by 1% and 0.3% for the
same energy ranges. A 1% variation in the stopping
power ratio, for air or TE-methane, is similar to the
uncertainties in the individual stopping power val-
ues themselves. Heavier charged particles produced
by proton interactions with the wall or gas material
also have stopping power ratios that are nearly
independent of particle energy if the particle speed
exceeds p = 0.05. Thus the mean value used is
insensitive to the proton energy spectrum above 10
MeV for air filling and above 1MeV for TE-methane
filling. Finally, as the range of protons in unit density
material a t energies below 1and 10 MeV is less than
30 pm and 1.2 mm, respectively (see Table 5.3 which
is also based on the ICRU values (ICRU, 1993)), any
variation of the stopping power ratio for those pro-
tons will affect only a small portion of the proton
energy distribution.
With respect to the uncertainty of the mean stop-
ping power ratio value, variations in the calculated
values in the most frequently used tabulations
(Andersen and Ziegler, 1977; Janni, 1982; ICRU,
1993) result in differences of up to 2% in ratios
needed for clinical proton dosimetry. The most recent
of these tabulations from the ICRU establishes
parameters that take into consideration available
data for all elements, compounds and mixtures. The
errors in the resulting stopping power ratios are
expected to be no more than 1% (1s.d.).
5.4.3.2 Considerations of w and W. Knowl-
edge of w or W is required for the conversion of
charge collected in an ionization chamber to depos-
ited energy. The value of W(E) for charged particles
create an electron-ion pair by an ionizing particle
-
which imparts all its energy to the gas. Wle has
- - -
units of JIC. A more understandable interpretation
of WfE) arises from recognizing that EIW(E) is the
mean number of ion pairs formed when particles of
energy E dissipate all their energy in the gas. Note
that W(E) or EIW(E) represents an average over all
For indirectly ionizing radiation such as photons
or neutrons, or when directly ionizing particles dissi-
pate all their energy in the gas, W(E) is the correct
dosimetric conversion coefficient from ionization to
energy imparted to the gas. Secondary particles of all
energies less than that of the indirectly ionizing
radiation are generated and interact with the gas
and make W(E) the proper choice (Verhey and Ly-
man, 1992). As protons, even for range modulated
beams, lose only a fraction of their energy in travers-
ing the gas, the proper conversion coefficient is the
differential value, w(E). Here AEIw(E) is the mean
number of ion pairs formed when a particle of energy
E expends AE in the gas. Use of w is appropriate
even for protons of as little as 500 keV energy, as
their range in air or TE-methane a t standard tem-
perature and pressure is more than one cm, signifi-
cantly larger than the dimensions of a typical ioniza-
tion chamber. As discussed below, w(E) and W(E) are
largely independent of energy for particle speeds
which are well in excess of orbital electron speeds.
When w(E) is constant, W = w.
The reasons for the variation of the value of W(E)
with particle energy and species are subtle and
uncertain. No explicit treatment of this problem is
available. A comprehensive discussion of these phe-
nomena is given in ICRU Report 31 (ICRU, 1979).
Generally, variations in W(E) or w(E) values occur
when the proton speed is similar to that of orbital
electrons in the stopping gas resulting in a competi-
tion between ionization and excitation as the protons
slow down. Above this speed, variations are reduced.
Figure 5.7, taken from ICRU Report 31, shows a plot
of measured W values for protons of different ener-
gies stopping in TE-methane gas. Similar data are
shown in Figure 5.8 for nitrogen. The variation in W
a t lower speed is apparent. Variations in W for alpha
particles and heavier ions of similar energies are
more significant due to their lower speed.
The ICRU (ICRU, 1979) recommends W l e values
of 31.0 t 1.5 JIC and 36.5 2 1.5 JIC for TE-methane
and nitrogen, respectively, for the proton energy
range 10 < E(keV) < 4000. As indicated by these
recommendations, few measurements of W or w are
awilable for protons a t higher energies. For air, the
ICRU suggests a value of 35.3 JIC below 1.8 MeV
proton energy and 35.2 JlC, the value determined for
5 MeV alpha particles, a t higher proton energies.
Bakker and Segre (Bakker and Segrit, 1951)
TABLE5.2-Proton mass electronic stopping power values for several elements and compounds taken from ICRU Report 49 (ICRU, 199.3~
,511, [MeV cm2 g-'I
E [MeV] A-150 Air Bone LiF Lucite Muscle Polyethy Polysty Teflon TE-Meth Te-Prop Water C A1 Si
 TABLE 5.3-Proton csda ranges for several elements and compounds interpolated from the tubulations of ICRU Report 49 IICRU, 1993)
E[M~V] A-150 Air Bone LiF Lucite Muecl
measured w value ratios for several gases a t 340-
MeV proton energy. They measured the w value
ratios of nitrogen and air to argon of 1.31 and 1.30,
respectively. Using ICRU Report 49 stopping power
values, these ratios would become 1.30 and 1.29,
respectively. Combining the recommended proton
Wle value for argon of 26.66 J/C (ICRU, 1979) with
the experimental ratios would then predict wle
values of 34.7 and 34.4 JIC for nitrogen and air,
respectively. Note that the W value for noble gases
such as argon is expected to be independent of
particle species.
Petti et al. (1986) measured w values for nitrogen
and argon at 150 MeV proton energy by ionization
yield. Proton energy loss was directly measured and,
therefore, no stopping power data were needed. They
report wle values for nitrogen and argon of 36.3 2
0.8 and 26.5 2 0.6 JlC, respectively. The argon and
Range Idm21
Polyeth Polyst Teflon TE-Meth TE-Prop Watrr C Al 91
nitrogen values are in good agreement with the
ICRU recommended values, based on low energy
proton measurements. However, the nitrogen value
is substantially larger than that derived from Bak-
ker and Segre's data. Applying Bakker and Segre's
measured air to argon w value ratio to Petti's argon
measurement, a wle value for air of 34.2 2 0.8 JIC is
predicted.
Hiraoka et al. have reported relative w values for
air, TE-methane, N2, C02, Ar, and methane for
electrons from 60Coy-ray interactions, 70-MeV pro-
tons, 43-MeV deuterons, 3He particles and neutrons
produced by charged particles (Hiraoka et al., 1988).
The neutron beam was created by bombarding a
beryllium target with 30-MeV deuterons. Ratios of u1
values were determined by substituting different
gases into an ionization chamber exposed to constant
irradiation. The resulting data were then normal-
 T E - M e t h a n e Gas
0 Leonard and Boring 1973
A Chemtob et al. 1978
0 Kuhn and Werba 1978
X Rohrig und Colvett 1978
*Thoinas and Burke 1986
Larson 19.58
Proton Energy / keV
Fig. 5.7. Experimental W l e values for protons stopping in
TE-methane gas (from ICRU Report 31 (ICRU, 1979)).
ized to W/e values of 33.97 J/C (air, electrons, BIPM
(BIPM Bureau International des Poids et Mesures,
1985)),36.5 JIC (N2,protons, ICRU Report 31 (ICRU,
1979)), 31.0 J/C (TE-methane, neutrons, ICRU Re-
port 31), and 36.5 JIC (Na deuterons and 3He,
assumed) and are summarized in Table 5.4. Normal-
izing to the nitrogen value, the w values for the
different gases for proton bombardment were found
to agree with ICRU Report 31 values except for
methane. For comparison purposes, the ICRU recom-
mended values for protons are included in this table
(ICRU, 1979). Note that the w /e value which Hiraoka
et al. obtained for air, 35.3, is in agreement with the
ICRU value of 35.2, but greater than the value 34.3
recommended by the AAPM protocol for heavy
charged particle therapy (AAPM, 1986). The Euro-
pean Clinical Heavy Particle Dosimetry Group,
ECHED, adopted the ICRU recommended w value
for air (Vynckier et al., 1991; Vynckier et al., 1994) as
did the Japanese particle groups (Hayakawa and
Schechtman, 1988).
Denis et al. (Denis et al., 1990) measured a w l e
value for nitrogen of 36.8 5 0.5 J/C for 66 MeV
protons in agreement with the value of 36.3 t 0.8 for
150 MeV protons measured by Petti et al. (1986) and
with the recommended ICRU value for much lower
proton energies. Their corrected value for air, 35.6 f
0.61, is in agreement with the ICRU reported values,
albeit a t lower proton energies, and with the work of
Hiraoka et al. a t similar energies.
Siebers et al. compared absorbed dose to water
determined with a calorimeter to that determined by
an air filled ionization chamber to deduce w near the
entrance region in a 250 MeV proton beam (Siebers
et al., 1995) . Measurements were made in a water
slab at a depth of 100 mm where the mean proton
energy was 180 MeV. A water calorimeter of the
Schulz design (Schulz et al., 1987) was used to
determine absorbed dose to water in the region
surrounding the thermistor. Using a dummy calorirn-
eter to provide the same physical geometry, an
air-filled ionization chamber was exposed to the
identical proton fluence. By equating the absorbed
doses determined by each device, the wle value for
air is given by
a i r - cal (5.12)
- Dwater -Sair, water,
Qair
where D$&, is the absorbed dose to water deter-
mined by the calorimeter, mai, is the mass of air in
the ionization chamber, s ~ , ~ is, the~ , mean air to
water mass electronic stopping power ratio for pro-
tons of this energy and Qa, is the charge collected in
the cavity. Stopping power values were taken from
ICRU Report 49 (ICRU, 1993). Two techniques were
used to determine mi,, one based upon a 60Co air
kerma calibration and the other based upon an
exposure calibration. A third approach used the
same 60Cobeam to directly calibrate the ionization
chamber to absorbed dose in water using the water
calorimeter. Results are summarized in Table 5.5.
The w value for air which Siebers et al. have
derived is consistent with that derived by combining
the argon w measurement of Petti et al. with the air
to argon w ratio measured by Bakker and Segrg.
Note that in this determination, the measured re-
sponse in both the calorimeter and ionization cham-
ber is due to all particles which deposit energy in the
gas, including nuclear secondaries, thereby closely
mimicking the situation encountered in clinical work.
Two further comparisons between ionometry and
calorimetry have been reported which can be used to
infer w-values for protons. Delacroix et al. made
measurements in range modulated proton beams
with effective energies of 33 and 52 MeV and in an
unmodulated beam a t 186 MeV (Delacroix et al.,
1994), using an A-150 plastic calorimeter and an air
filled ionization chamber in a solid plastic slab. If the
measurements are interpreted using ICRU Report
49 stopping powers, a w l e value of 34.3 2 0.3 J/C
achieves agreement between the calorimetric and
ionometric determinations a t the effective proton
energies. In a similar comparison, Seuntjens et al.
actermined the absorbed dose to water by calorim-
etry and air ionometry in an 85-MeV range modu-
lated proton beam (Seuntjens et al., 1994). These
results were also consistent with a wle value of
34.3 % 0.3 JIC.
 Nitrogen G a s
50 -- ' ' '
I " ' "'
' ' ' I ' "' I " -

0 Boring et al. 1965 A Nguyen et al. 1980 -

- + Lowry and Miller 1958 $ Thomas and Burke 1986 -
48 - $chalbr et al. 1963 + Denis et al. 1990 -
- Larson 1958 A Petti et al. 1986 -
46 '1
X Parks et al. 1972
0 Kuhnand Werbu 1978
0 Bakker and SegrL. 1951 -
-

-
32 I I I I l l

lo" I b6
P r o t o n Energy / keV
Fig.5.8. Experimental W l e values for protons stopping in nitrogen gas (from ICRU Report 31 (ICRU, 1979)).

The w value in air is not currently well established
for proton eaergies encountered in radiotherapy. All
low energy W(E)Ie measurements (ICRU, 1979) and
two w(E)le determinations near 65-MeV proton
energy, (Denis et al., 1990) and (Hiraoka et al., 1988),
are consistent with a value of 35.3 2 0.4 JIC.
Determinations of w l e for energies a t or above 150
MeV proton energy yield a somewhat lower average
value of 34.4 2 0.4 J/C. Using stopping powers from
ICRU Report 49, the most recent measurements
cited indicate a w value near 34.4. Clearly, more
comprehensive w value determinations are needed
in the energy range from 20- to 150-MeV proton
energy to resolve these differences.
The comparison of calorimetric and ionometric
measurements a t the same point in a proton beam
are particularly important as this procedure closely
mimics the clinical situation. Such measurements
inherently determine the effective product of the w
value and the mass electronic stopping power ratio of
water to air for the range of energies and particle
types present in the treatment beam, thereby avoid-
ing direct dependence on knowledge of w(E)le.
Determinations of w(E)l e and W(E)I e for nitrogen
and air are summarized in Table 5.6 and Figures 5.8
and 5.9. Until more accurate measurements are
available between 20 and 150 MeV, use of a w l e
value in air of 34.8 t 0.7 JIC for effective proton
energies above 50 MeV would be consistent with
calorimetric comparisons and with the available
direct measurements.
5.4.4 Determination of Absorbed
Dose to Water
The determination of absorbed dose in a phantom
using ionization chamber response depends on a
calibration of that response in reference conditions.
TABLE5.4-Estimated w Ie fprotons, deuterons and 3He)and W l e TAELE
5.6-Proton w l e and W l e values for nitrogen and air
(neutrons and 60Co)values in JC-' for the stated radiations. above 1.5 MeVenergy
Energies shown for 6OCo and neutrons are mean values. Values
Energy wle or Nitrogen Air
denoted by [f] are normalizing values for that radiation modality. (MeV) Wle (JICJ (JIG) Reference
Data were taken from Hiraoka et al. (Hiraoka et al., 1988). The
ICRU-proton values are from ICRU Report 31 (ICRU, 1979) 1.83 Wle 36.8 f 0.34 35.18 f 0.42 Larson, 1958
except for TE-methane which is taken from Goodman and Coyne 2.51 Wle 35.2 + 0.17 Thomas and
(Goodman and Co.yne; 1980) Burke. 1986
3.6 wle 36.6 f 0.7 Parks et al., 1972
Radiation aualitv
65 wle 36.8 f 0.32 35.6 f 0.61 Denis et al., 1990
Yh Neutrons
ICRU 0.4MeV Protons Deuterons JHe 3MeV
68.2 wle 35.3 t 0.7 Hiraoka et al.,
Ga5 protons electrons 70 MeV 35 MeV 95 MeV protons 1988
33-186 wle 34.3 f 0.4 Delacroix et al.,
Air 35.2 33.971 35.3 35.6 35.7 35.4 1994
TE-methane 29.3 29.4 30.4 30.9 31.0 31.0t 55 wle 34.3 + 0.4 Seuntjens et al.,
N2 36.5 35.1 36.51 36.5t 36.5t 36.2 1994
coz 34.4 33.4 34.3 34.7 34.9 35.2 150 wle 36.3 + 0.8 34.2a Petti et al., 1986
Ar 27.0 26.0 27.0 27.0 27.4 30.3 180 wle 34.4 t 0.4 Siebers et al., 1995
Methane 30.5 26.6 27.9 28.2 28.3 28.2 340 wle 34.7b 34.4b Bakker and Segre,
1951
" This value was deduced from the w(Ar),le value by using the
wle ratio of air to argon of Bakker and Segri: (Bakker and Segre,
There are three distinct methods for obtaining that 1951).
calibration which will be discussed in this section. Measured wle ratio values were deduced after correcting to
Detailed recommendations for the factors to use in ICRU Report 49 stopping powers and using the ICRU recom-
these calibrations will be discussed in Section 7. mended value for w (Ar)Je of 26.6 JIC for normalization.
5.4.4.1 Determinations Based on Fluence.
This section deals with the method of determining
absorbed dose beginning with a fluence measure- cup per unit charge (monitor unit) in the transmis-
ment of monoenergetic protons. As discussed in sion ionization chamber can be determined. The
Section 3, a fluence measurement leads to a determi- cema in water per monitor unit is then given by:
nation of cema. We will assume that the absorbed
dose in water will be equal to cema. Although there
are a number of devices which might be used for the where QE(E)is the fluence of protons of energy E
determination of fluence, such as scintillators or per monitor unit at the point in water
induction coils, we will assume the use of a Faraday where the cema is determined (15% uncer-
cup due to its simplicity and availability. The Fara- tainty)
day cup calibration technique in a proton beam (S(EIp), is the mass electronic stopping
normally starts with the implementation of a power in water for protons of this energy in
monoenergetic beam small enough in cross-section
to be completely accepted by the aperture of the cup.
Since this beam will not normally be used for treat-
ments, the cema which is determined at the location
of the cup must be transferred to a secondary device,
normally an ionization chamber which is small
enough to be irradiated uniformly in the calibration
beam (Verheyet al., 1979).A transmission ionization
chamber is used to monitor the total proton flux. If
the entire beam is accepted into the Faraday cup,
then the number of protons measured in the Faraday

TABLE 6.5- Proton w l e values i n air derived from calorimeter

determined dose values at 180 MeVproton energy. Value marked
with an asterisk [*I used a 6oCo-derivedcalibration of the
ionization chamber to absorbed dose in water using the
calorimeter (Siebers et al., 1995)
Calb~ratioamethod ude (JIG)
Proton energy / MeV
NK- air kerma 34.4 f 0.6
Fig. 5.9. Experimental W l e or w l e values for protons stop-
N,-exposure in air :34.6 +- 0.8
ping in air. Data are from Table 5.6. The line shown represents a
N,"-absorbed dose in water 34.2 ? 0.4
best straight line fit to the data. Errors are as given by the authors
Average 34.4 ? 0.4 and are not necessarily defined in a uniform manner.
 MeV cm2gL1(1% uncertainty (1s.d.) for a
well-defined energy) and
Cw is expressed in Gylmonitor unit.
In the case where the calibration beam is monoener-
getic, this can be rewritten as
where N is the number of protons of energy E
entering the Faraday cup per monitor unit and a is
the effective area of the beam in cm%ssuming
uniformity (1%-2%uncertainty). Mass electronic stop-
ping powers as a function of energy are given in
Table 5.2.
Once the cema in water at the location of the
Faraday cup is determined, a transfer ionization
chamber is irradiated a t the same point in the same
beam and its response determined per monitor unit
of the transmission chamber. This can be converted
into a response per unit absorbed dose as follows:
where J, = Dl(w(E)le) is the response of the trans-
fer ionization chamber per monitor unit in units of
collected charge per unit mass. In essence, this
constitutes a determination of (S(E)lp)Jw(E)le,result-
ing in a calibration of the transfer chamber in terms
of absorbed dose to the filling gas, under the gener-
ally accepted justification that cema and absorbed
dose are equal.
Knowledge of the beam area extends the Faraday
cup response to a fluence measurement. Due to the
simplicity of operation, the precision of a Faraday
cup fluence measurement is high. The accuracy of
the resulting dose calibration of the monitor cham-
ber in proton absorbed dose per unit collected charge
by this method is critically dependent on the knowl-
edge of the energy spectrum of protons in the calibrat-
ing beam as well as the effective area of the beam.
The assumption that the calibration of the transfer
chamber can be used in any other proton beam
without modification depends on the linear responbe
of the transfer ionization chamber with deposited
energy over the entire range of proton energies
present in the beam.
5.4.4.2 Determinations Based on Air Kerma
Calibrations This method is based on the determi-
nation of the mass of gas in an ionization chamber
using knowledge of the air kerma calibration factor
determined in a 'j°Co beam. When the gas in an
ionization chamber receives an absorbed dose, D , ,
from protons, the charge, produced Q,,, is
where m,is the mass of the gas, and w , ,the average
energy expended by protons in the gas to produce an
ion pair. The simplest method of determining the
mass m uses a known calibration factor for the
chamber for another radiation; in particular, NK,for
'j°Co gamma rays, which is obtained from a stan-
dards laboratory. This G°Co calibration factor is
combined with fundamental parameters and factors
which depend only on the geometry and material of
the chamber to determine the mass of gas in the
detecting volume of the chamber. That is:
where
and where NK is the air kerma calibration factor
for "Co in units of kermaheading
(independent of gas in chamber),
g is the fraction of secondary electron
energy lost to bremsstrahlung (ap-
proximately 0.003 (Boutillon, 1987),
s,&l, is the m e m ratio of restricted
mass stopping powers from wall ma-
terial to the gas for the secondary
electrons,
( / J ~ ~ I ~is) the
, ~ ,mass
~ ~ ~energy-absorp-
~
tion coefficient ratio from air to wall
for 'j°Co photons
AWal1 is a factor which corrects for the
absorption and scatter in the wall
and build-up cap (referred to as k,
in many protocols (ICRU, 1984a;IAEA,
1987),
AiOncorrects for ion recombination dur-
ing the calibration,
Khumcorrects for the difference in re-
sponse between ambient air and dry
air and
WJe) is the mean energy required to form
an ion pair in the chamber gas (nor-
mally ambient air) for photons in
units of JIC. When using ambient air, a
value of W,le should be used which
accounts for the presence of water vapor.
Note that N D ,is the absorbed dose to the gas in the
chamber in units of Gy C-l and is identical to Ng,
defined by AAPM (1983) and ND defined previously
by the ICRU (ICRU, 1984b). Note also that this
definition of N D ,assumes a homogeneous chamber
with wall and cap of the same material. Finally, it
should be noted that it assumes that ion recombina-
tion effects in the calibration condition are sepa-
rately evaluated and accounted for in Aion.In the
case that the calibration factor is based on exposure,
one can obtain NK&omNxby the expression NK(1-&) =

tion fador in units of R C-l. For ionization chambers
commonly used in clinics, recommended values for the
above fadors can be found in the literature (Nath and
Schulz, 1981;AAPM,1983;M A , 1987).Specificrecom-
mendations for commonly used ionization chambers are
given in Section 7.
Once the mass of gas has been determined by this
technique, the ionization chamber can be used to
determine the absorbed dose to water in a proton
beam as:
where Q , , is the charge produced in the chamber,
WE) is the proton fluence,
w,,(E)le is the mean energy required for
protons to produce a n ion pair in the cham-
ber gas (in JIC)
and s,JE) is the mass electronic stopping
power ratio of water to gas for protons of
this energy
The quantities s,JE) and w,,(E) are shown to
explicitly depend on energy. Since air is the recom-
mended gas for ionization measurements and since
w & , ~is assumed constant above 1MeV (ICRU, 1979)
and s , , is constant to within 3% above 1MeV and to
within 1% above 10 MeV (ICRU, 19931, the integral
is usually replaced by a constant value correspond-
ing to the effective energy of the beam. This will be
discussed further in Section 7.
5.4.4.3 Determinations Based on Absolute
Dose to Water Calibrations. With the increasing
availability and use of water calorimeters a t stan-
dards laboratories, it is anticipated that direct dose
to water calibrations in 60Co may soon become the
preferred method of ionization chamber calibration.
The formalism for use of such a direct absorbed dose
to water calibration factor is described in the litera-
ture (Hohlfield, 1988; Rogers, 1992; Rogers et al.,
1994; Medin et al., 1995). The advantage of a direct
dose to water calibration is that it eliminates all
uncertainties related to chamber-specific correction
factors such as those shown in Equation 5.18, as long
as the chamber is used in the same quality beam as
the calibration beam. For a 60Co absorbed dose to
water calibration, the dose to water in another @jCo
beam is simply
where My" is the meter reading corrected only for
differences between the current clinical use and the
calibration condition in parameters such as tempera-
ture, pressure and humidity and ND,,,,is the calibra-
tion constant in units of absorbed dose to water per
meter reading. For use in a beam of another quality,
such as protons, the dose can be obtained by
where
and where q"is corrected for differences in cham-
ber response in the 60Cocalibration beam relative to
the properties of the proton beam including ion
recombination and polarity effects. The factor k, is
very similar to the kQfactor introduced by Vatnitsky
et al. (1996). Note that all other chamber-sp&c
factors which are needed in a n absorbed dose to
water calibration are included in ND,,,,.
In the ideal case where a water calorimeter is used
to directly calibrate an ionization chamber in a
proton beam, only differences between the clinical
beam and the calibration beam would need to be
accounted for, namely,
and
and where p d refers to the proton calibration beam
and ( w ~ ) is
, assumed to be constant with energy.
 6. Beam Monitoring and Relative Dosimetry
To successfully exploit the high level of precision
potentially achievable with protons, the position and
of the beam relative to the patient
target must be known a t all times throughout the
treatment. Specifically, this could require monitor-
ing of the patient position, the position of the target
within the patient, the integrated beam intensity,
the instantaneous beam positior,, the location of
beam shaping devices, the beam energy and the
beam shape. Methods of monitoring radiation-
related parameters are discussed in Sections 6.1 and
6.2. Of course, a program of quality assurance of the
physical aspects of proton beam treatment is essen-
tial and also involves beam monitoring and relative
dosimetry. Quality assurance is discussed in Appen-
dix A.
6.1 Beam Monitoring
For safety reasons, two completely independent
dose monitoring devices with regularly verified cali-
bration are needed, which give a signal proportional
to the patient dose. Ionization chambers may serve
this purpose. If the accelerator used is capable of
high intensities or bursts of protons, special care in
the design of the monitor system is required. If the
dose monitors can be saturated by excessively high
dose rate, an additional less sensitive monitor is
necessary, which is capable of rapidly terminating
beam exposure. As an example, secondary emission
monitors (SEMs) may be used for this purpose. The
intensity monitors must be directly wired into the
interlock system using a fail-safe method which does
not allow operation of the beam if they are not
working properly.
For static beam delivery systems, beam centering
devices are required which are able to detect mis-
alignment between the beam and the beam modify-
ing elements, interrupting treatment if necessary.
For dynamic treatment systems, parameters such
as range shifter position, patient couch position, etc.,
which are not directly monitored by radiation detec-
tors, must be checked by continuous reading of
position through a channel which is independent of
the control unit for that element. A suitable element
may be an absolute angle decoder, for example.
For dynamic beam delivery techniques, it is also
necessary to use devices capable of monitoring the
energy and position of the beam a t all times. Such
devices need to be both fast and yield reproducible
results. In general, this will be done with a position
sensitive monitor just in front of the patient. Arrays
of parallel plate ionization chambers or multiwire
chambers are well suited for this purpose. Checking
the depth of penetration of protons is more difficult.
Indirectly, it can be verified by checking the position
of wedges or range shifter plates if the range is
controlled mechanically. The information must be
collected and processed rapidly enough to minimize
the possibility of mistreatment.
Detection of positron emitters produced within the
geometrical limits of the proton beam can help to
verify the position of the beam relative to the patient
and, to some extent, the beam penetration. The
ability to perform such measurements requires the
availability of a positron camera in the vicinity, due
to the short half-lives of the radionuclides produced.
The induced radioactivity can depend on the tissue
density and proton energy spectrum, as well as on
the absorbed dose a t the point of interest. A positron
camera within the treatment room could, in prin-
ciple, be used as a monitor of both the beam position
and beam penetration during the irradiation (Smith
et al., 1977).
Depending on the location of the target volume
within the patient, continuous monitoring of the
position of the patient andfor target volume could be
required. In extreme situations, such as irradiation
of lung tumors or targets within uncooperative pa-
tients, gating of the beam delivery system with
physiologically derived signals from the patients
(eg., breathing) might be desirable.
6.1.1 Detectors for Monitoring Beam
Intensity
Parallel plate ionization chambers are most fre-
quently selected as integral dose monitors. Air or
nitrogen filled chambers are typical choices. For
unsealed chambers, automated correction for atmo-
spheric pressure and temperature is recommended.
Multi-element ionization chambers can be used for
monitoring both beam intensity and beam profile.
The latter application will be' discussed in more
detail in Section 6.1.2. For scanning beam measure-
ments, a small air gap and high field strength of a
few kV/cm can yield response times less than 100 p
(Lin et al., 1994). Fast recycling charge integrators
for multi-channel readout are also available (Renner
et al., 1989).Wire chambers filled with gas that is not
electronegative can be used for the same purpose, as
they are well-suited for low dose rates and high
resolution if operated in the proportional mode.
6.1.1.1 Ionization Chambers for Beam Moni-
toring. Depending upon experimental conditions,
ionization chamber response can be related not only
to absorbed dose, but also to cema, current, fluence,
or fluence rate. For example, when the change in
proton energy during a cavity traversal can be
neglected, the average charge produced per proton
traversing the gas, Q INJE), is given by
where (S(E,)Ip), is the mass electronic stopping
power for the gas g at proton energy E,, p is the
density, xis the mean path length in gas, and w(EJ is
the mean energy needed to produce an ion pair for
protons of energy E, losing (S(EJ, . x) energy in the
gas. This constitutes a measurement proportional to
the proton fluence. A transmission ionization cham-
ber (TIC) is well suited to fluence or fluence rate
determination. In this device, one or more parallel
plate ionization chambers are positioned in series,
with the plates perpendicular to the proton beam
axis. Assuming that ionization in the TIC is due only
to electronic interactions of the primary beam pro-
tons, the ionization-per-proton monotonically de-
creases by a factor of about 2.5 for protons from 70 to
250 MeV. Figure 6.1 plots values of the ionization
produced per proton versus proton energy for air, He,
and Ar gas filling relative to the value a t proton
energy Eo = 250 MeV based upon:
This simple dependence on stopping powers shows
Proton Energy / MeV
Fig. 6.1. Values of the ratio of the charge per proton produced
by traversal of a 1 atm-cm gas path versus proton energy for He,
Ar and air filling relative to the value a t 250-MeV proton energy.
Values for He and Ar are multiplied by I .1 and 1.2, respectively, to
allow distinction amongst the plots.
that a calibration of the TIC at a reference energy, Eo,
predicts the response a t any other energy E,. Note
that this equation ignores any dependence of the
differential value w on proton energy. Mass elec-
tronic stopping power values for the different filling
gases are interpolated from the tabulations of ICRU
Report 49 (ICRU, 1993).
The usefulness of a TIC for relative fluence or
fluence rate determinations is apparent. As most of
the gas ionization is produced by primary protons,
the TIC walls can be of arbitrary material and of low
mass. Parallel plate geometry simplifies the elec-
trode design and yields a uniform electric field.
Electric fields in excess of 500 V mm-I may be needed
to ensure rapid charge collection with minimum
recombination effects.
Despite these advantages, there are some areas of
concern. For example, a t such large electric field
strengths, some deflection of the collection plates
will occur and produce an active chamber thickness
different from the assumed geometry. This will be
more important if thin foils are used for the chamber
plates. However, only volume changes that vary in
time or are unaccounted for are important.
If a noble filling gas is used, gas purity becomes
important (Jesse and Sadauskis, 1952; Jesse and
Sadauskis, 1953; ICRU, 1979). The so-called Jesse
effect results from small concentrations of impurities
in the noble filling gas yielding W values substan-
tially different from that of the pure gas. This is due
to stable excited states of the noble gas producing
ionization in the contaminant gas by energy trans-
fer. Figure 6.2 illustrates this phenomenon for He
filling gas and several contaminants. Such W varia-
tions are dependent on concentration which may be
time-dependent.
Due to the large interaction coefficients of the
primary protons, energy loss in a TIC can be mini-
mized by using small electrode gaps and gas pres-
sures below an atmosphere. However, gas pressure
stability and purity can then be problematic. Media-
tion of the uncertainty in ionization due to the Jesse
effect is possible by purposely adding a sufficient
concentration of a selected contaminant. Besides
reducing the uncertainty associated with the Jesse
effect, the contaminant produces larger electron drift
velocities enabling an improved TIC response time
(Bortner and Hurst, 1953).
Another problem may be recombination losses due
to the production of a large charge density produced
by high proton fluence rates. For example, a 250
MeV proton beam of 10 nA and 5 mm diameter,
crossing a 5 mm air-filled collection region produces
an ion current of 0.6 &and a n ion current density of
6 ~ A m r n a- t~atmospheric filling pressure. According
to Boag's theory of general recombination (Boag,
19661, the estimated ion collection efficiency for this
situation is about 80% for an electric field strength of
 1.6 _ ~ I ~ I I I ~ I ~ I I I I I I I I
- H e F i l l i n g Gas
-

0dr
* co, Area = 10 mm"
Cap = 5 mm

Electric Field / kV mm-'

0 2 4 6 8 10 12 14 16 18 Fig. 6.4. Fraction of ions collected in air versw collection
I m p u r i t y / Parts-per- 10,000 electric field for several proton currents calculated from Boag's
Fig. 6.2. Illustration of the Jesse effect in helium. Relative theory (Boag, 1966).
ionization is plotted as a function of concentration for various
kinds of impurities.

the ion collection fraction versus electric field strength

0.2 kV mm-l. Figure 6.3 shows a plot of the ion
collection fraction, f , versus beam area, A, for several
proton currents and for a 1kV mm-I electric field in
a 5 mm collection region filled with air a t atmo-
spheric pressure. Figure 6.4 shows a similar plot of
Beam A r e a / mma
Fig. 6.3. Fraction of ions collected in air versus proton beam
area for several proton currents, calculated from Boag's theory
(Boag, 1966).
for several proton currents, a beam diameter of 10
mm and a 5 mm gap filled to atmospheric pressure
with air. For a current of 10 nA, small area beams of
about 1011protons per second require careful design
of the TIC. Employing non-electronegative gases,
such as He or Ar, and lower pressures significantly
reduces the recombination problem with a concomi-
tant increase in dBculty of design and quality
assurance of stability.
An example of these ion recombination effects is
given by data collected during the commissioning
studies for the proton therapy facility constructed a t
Fermilab for Lorna Linda University Medical Center
(Cole et al., 1987). In this case, a TIC with a 6 mm
electrode gap filled to one atmosphere with air was
bombarded by different fluence rates of 230 MeV
protons. The electric field strength was 0.135 kV
mm-l. Measured ion collection fraction values are
plotted versus time-averaged proton fluence in Fig-
ure 6.5. Ion collection fraction values based upon
Boag's formulation assuming continuous irradiation,
are also indicated (Boag, 1966).The measured values
deviate significantly from Boag's constant current
prediction due to the time structure of the extracted
beam (Siebers, 1990). Specifically, the beam was
extracted non-uniformly over a one second interval
in bursts of less than 100 ps. These beam extraction
conditions produce large instantaneous proton flu-
ence rates and correspondingly low ion collection
efficiencies. Another TIC with a 3 mm air-filled gap
and electric field strength of 0.835 kV mm-I was

0 an instrument may be advantageous for monitoring
0 Data o
0 Monitoring of Dynamic Beam Delivery
0.65 I I t I I I I I I Determination of the absorbed dose delivered in a
0 1 2 3 4 5
Proton Beam Intensity / p- 10'4-m-Z-s-'
Fig. 6.5. Fraction of ions collected in air versus proton fluence
rate in a pulsed proton beam (Cole et al., 1987) and Boag's theory
(Boag, 1966) for a proton beam of constant intensity. The mea-
sured values are for beam intensities averaged over a one second
time interval.
found to have negligible ion collection losses under
similar bombardment conditions (Siebers, 1990).
6.1.1.2 Other Beam Monitoring Detectors.
Another monitoring device which is particularly
well-suited for large proton fluence rates is an elec-
tron secondary-emission monitor (SEM) (Taufest
and Fechter, 1955). Such devices consist of one or
more thin metallic foils mounted in an ultra-high
vacuum enclosure. As the foil is traversed by pro-
tons, electrons are released, resulting in a net cur-
rent flow that provides the signal (Laulainen and
Bichsel, 1972). Typically, less than one electron per
foil will be generated for each passing proton.
Karzmark provides a useful discussion of the perfor-
mance of SEM7sas applied to fluence monitoring of
electrons (Karzmark, 1964). For electron beams,
Karzmark reports 0.035 electrons per foil a t 8 MeV.
The nature of the device avoids recombination losses.
Hence the device has a large dynamic range with
excellent linearity. Several thin foils can be used to
improve the sensitivity with a small reduction in
proton range. The instrument can be calibrated for
different proton energies by observing the ratio of
response for several proton bombarding energies.
SEM's used in conjunction with TIC'S provide consid-
erable redundancy and a large dynamic range in
response.
Coutrakon et al. have tested a noble gas-filled
scintillator cell for purposes of proton beam intensity
monitoring (Coutrakon et al., 1991). A Xe gas path
length of 70 atm-mm was traversed by 230 MeV
protons. Gas scintillations resulted in a typical pho-
tomultiplier current of 25 mA for a nominal proton
current of 0.4nA. The observed signal-to-noise value
of 80 dB ensured accurate proton fluence rate deter-
mination with a large dynamic range. This device
could also be operated in single photon counting
mode for lower proton intensities, extending the
dynamic range even further. By detecting gas scintil-
lation, problems associated with ion recombination
present in ionization chambers are minimized. Such
proton fluence during active beam scanning due to
the large available dynamic range, good signal-to-
noise ratio, and immunity from saturation phenom-
ena.
6.1.2 Special Considerations for the
single treatment session must be performed i n much
the same way for dynamic or passive beam delivery
techniques. However, in dynamic beam delivery, the
readout of the dose monitor a t any time during the
treatment session has no simple correlation to the
total dose delivered to any portion of the target
volume. The reading of the monitor a t the conclusion
of the treatment contains information on the total
number of protons delivered, but not on their planar
distribution. Therefore, in addition to the total pm-
ton fluence, the fluence a t each location on the
surface of the monitor must be determined with two
independent measuring channels. This measure-
ment must be performed quickly and with a preci-
sion of better than 3% per location. Of t h e two
monitors, one is actively controlling the dose deliv-
ery while the other one, with a few percent higher
preset, is acting as a safety device.
The proton beam position must be verified to a
fraction of the beam diameter to avoid errors in the
overlapping of beam positions which could ulti-
mately create inhomogeneities in the dose distribu-
tion. This position measurement can be performed
with pad ionization chambers (Coutrakon et al..
1991),using pad sizes matched to the diameter ofthe
scanning beam. Pairs of strip ionization chambers
can serve the same purpose. To reduce the electron-
ics necessary for the readout of the elements, every
nth pad or strip can be connected to the same
readout. An even simpler technique is to use a
parallel plate ionization chamber with two measur-
ing electrodes and a zigzag high voltage electrode in
between (Bacher et al., 1989). The ratio of the two
ionization currents contains the information of where.
in relation to grooves of the high voltage electrode.
the beam has passed. A single tilted high voltage
plane can also give the position with respect to the
chamber, but with less precision (Bacher et al..
1989).
For line scanning, the beam profile along an entire
line must be compared with the desired profile. In
addition to the beam position on the entrance plane,
the range of the protons must be known (Bennett
andhehambeau, 1977).If a mechanical range shifter
is used, the verification of the material introduced in
the beam may be used as a check, once the correla-
tion with the actual range has been established.
For verifying the distribution of dose within the
treatment volume when using dynamic beam deliv-
ery, measurements in phantoms with density inhomo-
geneities or in anthropomorphic phantoms, are essen-
tial. Two examples of detectors which could be useful
for this are two dimensional detectors, e.g., track
detectors (Sunaga et al., 1988) or even 3-D gels of
FeS04,which are discussed in Section 6.2.1.3.
6.2 Dose Distributions
Relative dose measurements require no detector
calibration other than verification of appropriate
linearity of response within the assumed range of
measurement conditions. Relative dose measure-
ments are employed for routine daily clinical physics
activities, system commissioning, quality assurance,
research and development.
Measurements made during the commissioning of
a treatment system include mapping of clinical
radiation fields and determination of fundamental
beam characteristics, including beam range, unrnodu-
lated beam depth dose curves and dose distributions
within pencil scan beams. These characteristics are
necessary for design and control of the beam delivery
system. For example, unmodulated depth dose data
are used to design range modulators. Once the
delivery system is ready for patients, clinical beam
field mapping measurements are made for the range
of treatment conditions.
Proton patient portals require an individual physi-
cal calibration measurement or computation of treat-
ment dose monitor setting from a calibration model.
Individual patient calibration is performed with a
dosimetry system having a known dose calibration
relative to the proton facility primary dose standard.
Dosimetry data are acquired as relative dose mea-
surements during commissioning and from accumu-
lated individual patient calibration data.
6.2.1 Detectors for Dose Distribution
Measurements
Detectors employed for relative dosimetry must
have the appropriate sensitivity, energy indepen-
dence, response linearity and spatial resolution for
each clinical dosimetry task. Ionization chambers,
silicon diodes, x-ray film, diamond detectors, TLD,
and radiochromic film can be employed. Since dose
gradients a t the edges of proton beams are rather
steep, ionization chambers are often considered too
large to accurately map dose distributions in proton
tion.
The irradiation medium for clinical dosimetry
should simulate patient treatments as closely as
possible in terms of phantom composition, location
and extent. Water should be regarded as the stan-
dard reference medium. However, water-like substi-
tute materials may be used provided that the behav-
ior of the beam within the medium relative to its
behavior in water is understood.
The time structure of the beam must be considered
in designing clinical dosimetry measurements. A
detector must dwell a t the same location for many
beam cycles to obtain reproducible results. Ex-
amples of cyclic behavior which must be considered
include the accelerator duty cycle, time structure
within a treatment beam pulse, pulsed irradiation of
a rotating modulator wheel, and dynamic beam
spreading. Dosimetry measurement systems with
multiple detectors in linear, area or volume arrays
save time for measurements of dose distributions,
particularly in dynamic systems. It should also be
noted that many relative dose measurements em-
ploy a separate detector as a reference monitor to
correct for dose rate variations.
6.2.1.1 Silicon Diodes. Semiconductor diodes
have routinely been used for absorbed dose measure-
ments. Because of their small volume, typically 0.1
mm3or less, excellent spatial resolution is achieved
with good sensitivity. Due to these advantageous
features, Si diodes have been widely used in radia-
tion dosimetry (Gulbranden and Madsen, 1962; Raju,
1966; Koehler, 1967; Trump and Pinkerton, 1967;
Smith et al., 1977). In most cases, Si diodes are
operated without external bias, in the so-called
photovoltaic mode, where the intrinsic depletion
region is used to produce charge flow (Klevenhagen,
1977; Maruhashi, 1977).
As the charge flow is by impurity carrier in t h e
diode junction, large instantaneous doses or dose
rates, even of low-LET radiation, produce a non-
linear dose response in n-Si diodes (Rikner and
Grusell, 1987). Using p-Si diodes reduces this effect
(Grusell and Rikner, 1986). Additionally, due to
lattice damage (Knoll, 1989), the sensitivity per unit
absorbed dose varies with the magnitude of previous
exposure. Such lattice damage is dependent upon
the type of particle producing the defects, with
greater damage resulting from more massive par-
ticles. Rikner measured the relative damage from
equal doses of 8 MV x rays, 20 MeV electrons, and 70
MeV protons and established a relative damage ratio
of 1: 20 : 40, respectively (Rikner, 1983). For 70 MeV
proton bombardment, the damage from a 10 kGy
exposure reduces the sensitivity to 30% of the value
prior to irradiation.
When Si-diodes are used for absorbed dose determi-
nations for protons, an energy dependent response
 ifferent trom that ol lonlza~loncnarIluals, IB UIJ-
erved. Koehler (Koehler, 1967) and Raju (Raju,
966) observed a discrepancy between diode re-
ponse and gas ionization chamber response near
he Bragg peak region. These results were confirmed
~ymeasurements at Louvain-la-Neuve with both.
nonoenergetic and spread-out Bragg peak beams4.
pigure 6.6 shows a plot of diode and parallel plate
onization chamber response versus proton residual
.ange for a monochromatic beam. Values are normal-
zed to the Si-diode peak response. The ionization
:hamber response is about 93% of the diode response
tt the Bragg peak. Note that this difference is not
?xplained by differences in stopping power between
2ir and Si which would produce a correction increas-
~ n gthe difference. Columnar recombination may
:ontribute to the observed effects.
Case et al., (Case et al., 1994) modeled the re-
sponse of thimble and parallel plate ionization cham-
bers and silicon diodes using a stochastic proton
transport technique. They demonstrated geometry-
induced effects in response in the thimble ionization
chamber which they claim might explain part of the
difference between the response of diodes and thimble
zhambers near the distal portion of the Bragg peak.
3eometric effects would not explain the difference in
response between plane parallel ionization cham-
3ers and silicon diodes.
Whatever the explanation, silicon diodes are clearly
not giving a response that is similar to absorbed dose
in tissue at proton energies below 20 MeV as mea-
sured by ionization chambers. To interpret their
response as absorbed dose in tissue requires knowl-
edge of the proton energy fluence spectrum at each
point of measurement, particularly for protons with
energies below about 20 MeV. Hence, caution is
advised when using silicon diodes for proton dosim-
etry.
6.2.1.2 Films. Radiographic film can be a very
convenient method of measuring proton relative
doses. The response of film is based on the formation
of a latent image in microscopic silver halide crystals
(grains) dispersed uniformly on a gelatin base (emul-
sion). The development process reduces the affected
grains to silver, while the fixing process removes
unirradiated grains. The resulting optical density is
proportional to the fluence of particles passing
through the emulsion (Dudley, 1966). As a result,
changes in optical density across the film can only be
interpreted as changes in dose if the energy spec-
trum of the protons is constant. Therefore, films can
be conveniently and safely used to measure the
distributions perpendicular to the proton beam direc-
tion, but not along the beam axis to measure depth
dose distributions. In an intermediate situation,
Personal communication:S. Vyneckier, Universitk Catholic de
Louvain, Bruxelles (1993).
I
I
o I o n Chamber A
R a n g e - 5 mm A
AA 00
A 0
Reaidual Proton Range / mm
Fig. 6.6. Relative Si-diode and parallel plate ionization cham-
ber response as a function of residual range. Values are normal-
ized to the peak Si-diode response. The ionization chamber
response is displaced by 5 mm for ease of comparison. Adapted
from Vynckier. (Personal communication: S. Vynckier, Universib5
Catholique de Louvain, Bnurelles, Belgium.)
where film is to be used to verify the dose distribu-
tion of irregularly shaped volumes, and where there
is a mixture of protons with different energies, an
expected film density distribution should be calcu-
lated on the basis of the known energy distribution
and a measured density-fluence calibration curve of
the film and compared with the measured distribu-
tion. In spite of the fact that film does not give a
response which is linear with dose, the simplicity
and convenience of film makes it a very useful
medium for studying the changing fluence distribu-
tion of a proton beam as it passes through matter.
6.2.1.3 Other Detectors for Relative Dosim-
etry. Small tissue-like integrating dosimeters allow
a more precise determination of absorbed dose at
high spatial resolution than many other dosimeters.
The crystalline amino acid, L-alpha-alanine, is a
good example. Alanine is a solid hydrocarbon in
micro-crystalline form that, when bombarded by
ionizing radiation, produces free radical products
that can subsequently be quantified by electron spin
resonance (ESR) spectroscopy. Similar to TLD dosim-
eters, alanine acts as a single target system, exhibit-
ing a linear response with absorbed dose until satu-
ration occurs. Regulla and Deffner (Regulla and
Deffner, 1982) report a tissue-like absorbed dose
response to a wide range of photon energies. For 200
mg samples, they observed a linear absorbed dose
response range from 1 Gy to 105 Gy. As with other
solid state dosimeters, the absorbed dose response
depends upon LET, i.e., the spatial pattern of energy
depositions by ionizing radiation. Hansen and Olsen
(Hansen and Olsen, 1985) report integral ESR re-
sponse values relative to gamma-ray or electron
bombardment for 16 MeV and 6 MeV protons and 20
MeV alpha-particles of 1.00, 0.86, and 0.58, respec-
tively. For these bombardment conditions, the aver-
age mass electronic stopping powers in alanine were
38,119, and 534 MeV cm2g-l, respectively. While the
relative sensitivity per unit absorbed dose varied
with stopping power, the observed linearity and
saturation dependence was independent of stopping
power, that is the free radical production was depen-
dent upon stopping power as a simple scaling factor.
Low-LET radiation results in the production of free
radicals which are stable a t room temperature.
However, for high-LET radiation, significant fading
occurs which depends upon the total absorbed dose
delivered. Hansen and Olsen (Hansen and Olsen,
1989) reported 2.5%, 3.2% and 3.7% fading a t 1000 h
following exposure to 16 MeV, 6 MeV, and 1 MeV
protons. Even for low-LET radiation, Regulla and
D e f i e r (Regulla and DeEner, 1982) reported 10%
fading a t elevated storage temperatures of 70 OC, as
well as increased yields for irradiations at elevated
temperatures, 20% a t 90 OC and lo5 Gy. Use of the
alanine dosimeter for clinical proton beam absorbed
dose determinations requires consideration of the
proton energy fluence spectrum and control of the
environment during and subsequent to bombard-
ment.
The conversion of Fe2+ to Fe3+ in a n aqueous
ferrous sulfate solution by radiation can be mea-
sured by optical densitometry a t 305 nm, but also by
magnetic resonance. MRI of gels containing FeS04 is
being investigated by several authors (Gore et al.,
1984; Hazle et al., 1991; Podgorsak and Schreiner,
1992; Maryanski et al., 1994) as a potential 3-D dose
verification system.
Other detectors such as TLDs, diamond detectors
(Vatnitsky et al., 1995) and radiochromic film (Nich-
iporov et al., 1995) are being used as relative proton
dosimeters. Since their energy-dependent response
characteristics have not yet been well documented,
their use has been limited.
6.2.2 Determination of Dose Distributions
The description of the changing energy and spatial
characteristics of a proton beam as it passes through
matter is critical for predicting the dose deposition of
the protons. The relative dosimeters described above
can be very helpful in determining these characteris-
tics. The determination and parameterization of the
dose distribution in matter is important for the
programs.
6.2.2.1 Measurement of Beam Range and
Depth Dose Characteristics. A critical aspect of
proton beam treatment is the ability to stop the
beam a t a specific point within the patient. Accurate
control of the stopping point depends on knowledge
of the beam range in water and of the water-
equivalent path lengths (wepl) of materials placed in
the beam path and of the tissues traversed. Central
axis beam range in water is measured for the beam
energies and range absorbers which are to be used
clinically. Depth dose distributions should be deter-
mined for a selection of energy, range modulation,
field size and other treatment parameters that may
affect the distribution. The slope of the distal fall-off
zone will depend on incident beam energy, energy
spread within the accelerated beam, and range
straggling produced by absorbers in the beam deliv-
ery system and by the patient.
Range and depth dose measurements may be
made directly in a water phantom or in a suitable
substitute medium. A suitable phantom should have
a n effective atomic number and density close to t h a t
of water to assure that both the scattering and range
properties are appropriate. Such a material may be
characterized by measuring its effective path length
relative to water. A solid phantom with characteris-
tics known to be similar to water has the advantage
of very accurate thickness or depth determination.
Phantoms made from polystyrene or other plastic,
with optional cavities machined for different detec-
tors, plus a selection of variable thickness absorber
slabs of the same material, can be useful.
Accurate depth dose measurements must be made
throughout the Bragg peak region, including the
sharp distal fall-off region near the end of the range.
This requires a detector with good spatial resolution
and with a dose response which is independent of the
variation of proton energy with depth. Although
ionization chambers and silicon diodes would appear
to be useful for depth dose measurements, caution is
called for when using silicon diodes since, as was
noted earlier, they have been observed to have
approximately 10% higher response than parallel
plate ionization chambers in the Bragg peak region
(Raju, 1966; Koehler, 1967; Vynckier, et al., 1994).
Radiographic film should not be used for depth dose
measurements because of its significant variation in
response per unit dose as a function of depth within a
proton beam.
Water-equivalent path lengths for materials placed
in the beam path are measured, and, for purposes of
treatment planning, the relationship between wepl
and CT number (based on relative x-ray attenuation
coefficients) for body tissues is established. The
distribution of dose as a function of depth, deter-
mined as a basic characteristic of the beam, is also
required data for treatment planning. The wepl for
materials may be measured by submerging samples
of known thickness in a water phantom and measur-
ing the effect on beam range. The wepl of materials
used for patient-specific devices such as range shifters
and tissue compensators, need to be determined. In
addition, the wepl of tissue substitute materials
should be determined and correlated with observed
CT numbers for the same materials.
The variation of depth dose characteristics with
field size should be measured. The shape of the depth
dose curve may differ significantly for small fields
and at locations within irregularly shaped fields
where the lateral extent is less than a few penumbra
widths. An understanding of this variation is impor-
tant for treatment planning and treatment field
calibration. Depth dose near the surface of treat-
ment fields may be affected by protons scattered
from field shaping apertures. Aperture scatter ef-
fects also may be more pronounced for small and
irregularly shaped fields.
6.2.2.2 Lateral Dose Uniformity and Beam
Penumbra. For proton treatment beams which are
produced by passive beam shaping (scattering) tech-
niques, the lateral uniformity of proton treatment
fields should be comparable or superior to that of
conventional photon and electron fields. Uniformity
may be expressed in terms of field symmetry for
points equidistant from the beam central axis and as
flatness variation over some designated portion (e.g.,
80%) of the field area. Uniformity characteristics
should be measured at several depths for the variety
of available treatment planning dose distributions.
The shape of the dose distribution at the lateral
field edge is extremely important in planning proton
beam treatments. Field placement in proximity to
radiation sensitive normal tissues depends on accu-
rate knowledge of the penumbra as well as consider-
ation of uncertainty in patrent alignment. Penumbra
widths may be defined as the distance separating
stated dose levels (e.g.,80%to 20% of the central axis
dose at that depth). Penumbra characteristics, which
can be determined by beam profile scans, will depend
on the design of the beam delivery system and will
vary with most treatment parameters including
beam energy, range shift, depth in the patient and
collimator-to-skindistance (Urie et al., 1986). These
variations should be accurately measured and repro-
duced within the treatment planning system.
Lateral uniformity measurements should be made
in water or in a water-equivalent phantom with a
detector having high spatial resolution in the scan
direction. Energy independence for lateral scanning
is not as important as for depth scanning. Silicon
diodes, diamond detectors and small ionization cham-
bers are useful. Radiographic film placed in a phan-
tom perpendicular to the beam direction provides
results similar to the other detectors. The film
should be scanned with a system which has good
spatial resolution and a film density response correc-
tion should be made.
 7. Recommendations for Determination of Absorbed Dose in a Phantom
7.1 General Recommendations
This Section presents recommendations for the
determination of absorbed dose in a .phantom ex-
posed to a clinical proton beam. These will include
the choice of dosimetric instruments, the methods of
use of these instruments and the choice of values to
be used for physical parameters needed to determine
absorbed dose a t a point in a phantom. These
recommendations are particularly important since
there are currently no internationally accepted stan-
dards for the dosimetry of proton beams. The need
for international uniformity of dose specification
between institutions continues to grow with the
number of clinical proton facilities. Existing regional
standards (AAPM, 1983; Vynckier et al., 1991;
Vynckier et al., 1994)cannot provide that uniformity.
7.1.1 Reference Dosimeter
The choice of reference dosimeter to be used for the
measurement of proton absorbed dose is determined
by the accuracy and sensitivity required as well a s by
size, convenience and availability, It is recommended
that thimble ionization chambers with a =OCocalibra-
tion factor, traceable to a standards laboratory, be
adopted as the reference dosimeter for proton beams.
The calibration factor can be based on exposure, air
kerma or absorbed dose to water. Although the
choice of chamber geometry and wall material are
not critical, a n open, ambient air-filled graphite or
A-150 tissue-equivalent ionization chamber is recom-
mended. This should ideally be of a standard con-
struction for which chamber-specific parameters are
known (AAPM, 1983; IAEA, 1987). These chambers
are generally available and their behavior in 'j°Co or
x-ray beams is well documented. For use in large
beams (L 5 cm diameter), chambers of 0.5 cm3 or
larger volume are recommended, for purposes of
stability. For smaller beams, chambers of approxi-
mately 0.1 cm3 volume can be used to improve
spatial resolution. The use of methane-based TE gas
(Rossi and Failla, 1956) in these chambers may be
useful as an additional confirmation of the dose
determination, since the stopping power ratio of TE
methane gas to water is close to unity (see Table 7.1).
In addition, when used in an A-150 chamber, the
atomic composition of TE methane gas is well
matched to that of the wall, thereby minimizing the
effects of any wall interactions of protons or other
particles which could result in ionizations in the gas.
Efforts should be made to confirm the accuracy of
the derived proton calibration factor of the reference
ionization chamber by using a calorimeter, prefer-
ably water-based. However, in the absence of calorim-
etry-based proton dosimetry standards supported by
standards laboratories, the recommendations of this
Report are aimed a t achieving uniformity of dose
determination between proton facilities worldwide.
7.1.2 Phantom Material and Reference Depth
The choice of phantom material is determined by
the choice of reference dosimeter. For an ionization
chamber, it is recommended that absorbed dose be
measured in water or other materials which are
close to tissue in electron density. When using mate-
rial other than water, the depth of measurement
should be scaled to the equivalent depth in water
using measured equivalent depths or the continuous
slowing-down approximation (csda) ranges from
ICRU Report 49 (ICRU, 1993). For calorimetric dose
determinations, thermal energy is measured in t h e
material of the calorimeter, converted to absorbed
dose in the calorimeter and then to absorbed dose in
the material of interest, which is recommended to be
water. Therefore, a water calorimeter would be t h e
most direct determination of absorbed dose to water.
For the Faraday cup, cema in water can be deter-
mined by a fluence determination in air followed by
a n ionization measurement in air.
For most clinical applications, a modulated spread-
out Bragg peak (SOBP) is required. The center of
such a SOBP is in a region of uniform dose. For this
reason, and for clinical relevance, a point near t h e
center of a SOBP is recommended a s the preferred
reference point for calibration.
7.2 Determination of Proton Absorbed Dose
to Water using a Calibrated Ionization
Chamber
As described in Section 5, standards laboratories
can provide calibrations of ionization chambers in
reference beams (normally 'j°Co) which can be used
to obtain a calibration factor relevant for use in
proton beams. These standard calibrations can be
based on exposure, air kerma or absorbed dose to
water.
7.2.1 NK-basedCalibrated Ionization
Chamber
As described in Section 5.4.4.2, if we assume that
and Wp can be considered independent of
(s,,,~~)~
energy, the absorbed dose to water, when using an
ionization chamber can be written as:
where M;" is the product of the meter reading 31
TABLE7.1 -Proton mass electronk stopping powers of various materials versus csda range i n water using values from ICRU Report 49
(ICR U, 1993)

E Water range S/p (MeV cm2 g- '1

(MeV) (dcm') Water A150 Air Muscle Polystr TE-Methane

and the corrections P,, (pressure, temperature), Pion
(ion recombination factor) and Pj, the product of all
other factors which can produce modified response
relative to the calibration condition,
N K (-
~g M w a 1 4 i o n
ND'g = s w d l , g ( ~ e n l ~ ) ~ r , w a l & h u m '
and where the definitions of the individual factors
are described in detail in Section 5.4.4. The value of
the quantity N D ,depends on the known properties of
the calibrating beam (usually 60Co) and the calibra-
tion factor of the chamber, provided by national
standards laboratories. The 60Co calibration factor
can be reported either as NK, a n air-kerma calibra-
tion factor, or Nx, an exposure calibration factor. The
relationship between the factors is:
N K ( l - g) = Nx(W,le)(2.58 x 10-*C/R-kg)), (7.2)
where g = 0.003 (Boutillon, 1987), is the fraction of
secondary electron energy lost to bremsstrahlung 7.3 Determination of Proton Absorbed Dose
and the units of Nx and NK are in WC and Jkg-C, to Water using a Calorimeter
respectively. Since N D , = (Wcle)lm,, one can see
Calorimeters provide the most direct method of
that a determination ofND,is equivalent to determi-
absorbed dose determination (see Section 5.3) and
nation of the effective mass of gas in the ionization
are an excellent choice for a primary standard.
chamber from which ions can be collected.
However, due to the fact that calorimeters are not
Recommended values for (wairIpand (sw,air),, as well
generally available at proton facilities and are more
as estimated uncertainties in their values, are dis-
difficult to use than ionization chambers, they have
cussed in Section 7.5 below. The uncertainty in the
not been chosen as reference dosimeters for proton
absorbed dose determined by ionization chamber
therapy. They are, however, candidates for reference
dosimetry is dominated by the uncertainty in ( w , ~ ~ ) ~
use at national and international standards laborato-
which is estimated to be 2 2%.
ries. Appropriate examples of calorimeters have been
A dosimetry worksheet for determining the ab-
constructed of A-150 tissue equivalent (TE) plastic
sorbed dose in water for a clinical proton beam, using
(Caumes et al., 1984; McDonald and Domen, 1986),
an ionization chamber calibrated in a 60Cobeam, is
graphite (McDonald, 1987),and water (Domen, 1980;
set out in Appendix B.
Schulz et al., 1987; Domen, 1994). When available,
It should be noted that Equation 7.1 assumes no
such calorimeters can be used to confirm the proton
contribution from the wall to the ionization in the
calibration factor of the reference ionization charn-
chamber gas. This is due to the fact that the mean
ber.
range of the secondary electrons from proton interac-
The calorimeter can be used to determine dose to
tions is so small that the vast majority of electrons
water in a proton beam as follows:
detected in the gas cavity are produced in the gas
(Laulainen and Bichsel, 1972). To the extent that D, = A T k (1+ TD)sW,,, (7.5)
small remaining wall effects can contribute to the
where AT is the temperature rise due t o radiation
ionization detected in the cavity, it is prudent to
(degrees Celsius),
choose ionization chambers with wall materials which
have stopping powers well matched to the reference
k is the specific heat of the sensitive element
in J kg-l ~c-l,
medium.
The effects of nuclear interactions are excluded in
TD is the thermal defect (or excess) due to
deposited energy which does not end up
the above description. Due to the fact that the ratios
as temperature increase. Note that a posi-
of stopping powers for charged particles are rela-
tive thermal defect means that some de-
tively independent of species and energy, the conver-
posited energy is lost in rearrangement of
sion of ionization in gas to dose in water is very
the lattice structure and the temperature
nearly independent of the effects of nuclear interac-
rise is therefore too small. For A-150
tions of the protons.
plastic, the thermal defect has been mea-
sured to be 4 t 1.5% (McDonald and
Goodman, 1982) and for water at 4 "Cti t
7.2.2 Nw-basedCalibrated Ionization is indirectly determined to be 0 2 1%
Chamber (Schulz et al. 1991). For graphite we
With a direct absorbed dose-to-water calibration assume 0% thermal defect.
in a 60Cobeam, the determination of absorbed dose to s , , is the ratio of mass stopping power of
water in a proton beam is simplified. In this case, as water to the calorimeter material (i 1%
shown in Section 5.4.4.3, the absorbed dose to water uncertainty).
in a proton beam can be written as: AT is normally determined with the help of a
Wheatstone bridge which is used to measure the
Dw,p = M r m N ~ , w , & p , (7.3) change in resistance of a thermistor in thermal
where M","" is defined as in Equation 7-1, contact with the calorimeter. The specific heat k is
ND,,, is provided directly by the standards labora- usually measured by passing an accurately known
current through a heating resistor which is in ther-
tory and is the dose to water in the 'j°Co calibration
mal contact with the calorimeter, and measuring the
beam in Gy C-I and
temperature change for a known amount of dissi-
(sw,air)p(wairIp pated energy. In the case of water, the specific heat is
k,=-- (7.4) well known. These techniques are discussed in a
(~w,air)c (wairlc '
number of references (Domen, 1986; McDonald and
as given in Equation 5.22. The recommended values Domen, 1986). In the case of A-150 plastic, the
of the quantities in Equation 7.4 are discussed in thermal defect has been measured by completely
Section 7.5. stopping very low energy protons in A-150 material
and measuring the temperature rise (McDonald and
Goodman, 1982).The mass electronic stopping pow-
ers needed have been calculated (ICRU, 1993) and
are energy dependent, but the ratio required for the
calorimetric dose determination in water is only
weakly dependent on proton energy for energies
above about 1MeV.
The uncertainties of the mass electronic stopping
power ratio and the thermal defect give a combined
uncertainty of about 2% to 3% in the calorimetric
dose determination in A-150 plastic, but in the case
of graphite or water, that uncertainty might be
somewhat reduced due to a smaller uncertainty in
thermal defect (Schulz et al., 1987), although that
conclusion is not universally accepted (Domen and
Lamperti, 1974; Domen, 1980).
7.4 Determination of Proton Absorbed Dose
to Water using a Faraday Cup
As a secondary standard, a Faraday cup or other
fluence-measuring device could be selected as indi-
cated in Section 5.2. This technique is more sensitive
to the energy distribution of the proton beam than
the ionization chamber or calorimeter and, there-
fore, caution is recommended when using fluence
determinations as a basis for clinical dosimetry.
As discussed in Section 5.4.4.1, a fluence measure-
ment in a proton beam of known geometry and
energy composition can be used to calibrate a trans-
fer ionization chamber for use in a clinical beam. The
Faraday cup is a convenient device for making such a
fluence measurement. If one assumes that a trans-
mission ionization chamber is used to monitor the
beam fluence, the dose to water can be obtained by
fluence measurement followed by a charge measure-
ment with an ionization chamber a t the same point.
If it is also assumed that the proton beam is monoen-
ergetic, then:
where N is the number of protons per monitor unit
in the Faraday cup,
a is the effective area of the beam in cm2,
assuming uniformity,
(Slp), is the electronic mass stopping power
of the protons of this energy in water in
MeV cm2g-l and
D, is expressed in Gylmonitor unit.
In the case where the calibrating beam is composed
of a mixture of proton energies, the mass electronic
stopping power in Equation 7.6 must be replaced
with an integral over the proton spectrum as indi-
cated in Equation 5.13.
With a carefully constructed Faraday cup and with
good beam geometry (Verhey et al., 19791,the fluence
of the proton beam can be determined to 1% or
better. The determination of the effective area of the
beam depends on the assumption that the protons
emanate from a single point. Methods for experimen-
tally determining the effective area have been de-
scribed (Verhey et al., 1979). The mass stopping
power for protons, necessary for the above conrer-
sion of fluence to dose, depends critically on an
accurately known beam energy. The presence of a
small admixture of low energy scattered protons can
lead to significant errors in absorbed dose determina-
tion (Verhey et al., 1979). Monte Carlo calculations
may be helpful in estimating the effect of these low
energy contaminants on the calibration. Amonoener-
getic proton beam constructed without collimators
has been reported to be capable of avoiding the
production of low energy scattered protons (Grusell
et al., 1995). The effect of nuclear interactions is to
increase the apparent mean deposited energy per
proton. This increases the predicted dose to water
per proton, depending on proton energy, by several
percent (Seltzer, 1993).The uncertainties introduced
by nuclear effects and by the sensitivity of the
calibration to the energy and identity of the beam
particles combine to make a fluence-based calibra-
tion potentially less accurate than other methods
discussed in this section.
7.5 Numerical Values of Required Quantities
The uniformity of approach as outlined must be
combined with an agreement on the numerical val-
ues of the physical parameters needed to obtain
absorbed dose to water from measured quantities.
For the mass of gas determination in the ionization
chamber, the factors needed, including (W,le), A,d
(Nath and Schulz, 1981), s ~and~( ~ , n
~ / p )~, wfor
U, a ~
specific chamber type in a 60Cobeam, can be calcu-
lated on the basis of chamber construction ( k Q M ,
1983; IAEA, 1987). For some commonly used graph-
ite and tissue-equivalent chambers, the recom-
mended values are listed in Table 7.2. For (W, 'e), a
value of 33.97 + 0.15 JIC for dry air has been
recommended based on a reevaluation of several
experiments using updated ionization potentials
(Boutillon and Perroche-Row, 1987). This value
should be reduced to 33.77 JIC to account for typical
water vapor content if ambient air is used in the
chamber (ICRU, 1979; Schulz et al., 1986). This
recommendation assumes that the calibration done
at the standards laboratory used ambient laboratory
air and did not correct the resulting calibration
factor to dry air. If the calibration factor is corrected
to dry air, then subsequent readings should be also
corrected to dry air and a value of 33.97 JIC for W,le
should be used. When using ambient air, a correction
for the response in humidified air compared to dry
air should also be made, as represented in the factor
Khumin Equation 7.1. A value of 0.997 is recom-
mended for Khum(Schulz et al., 1986).An additional
 TULE 7.2 -Recommended values ofphysical parameters for 0.2%. For ( w,,,),le , a constant value of 34.8 JIC is
common cylindrical ionization chamhers for 60Cobeams ( N a t h recommended since the range of protons of 1 MeI:
and Schulz, 1981;AAPM,1983; Gastorfet al., 1986;ZAEA, 1987)
- below which wle appears to be increasing (ICRU,
Chamber Demption Wall I ~,,II,~,, (R,/P),,~,,~,II 1979),is less than 25 pm in water.
Capintec, 0.1 cm3 C 0.991 1.010 0.999 For calorimetry, required physical parameters are
Exradin 0.5 cm3Spokas-T2 TE 0.985 1.145 0.906 the specific heat and the thermal defect. The specific
Exradin 0.05 cm" heat can be considered an experimentally deter-
Shonka-TI TE 0.992 1.145 0.906 mined quantity. In the case of water, the specific heat
NEL Farmer 0.6 ad-2571 C 0.990 1.010 0.999
is known to be approximately equal to 1.000 cal g-I
NEL Farmer 0.6 cm3-2581 TE 0.990 1.145 0.906
Far West 1.0 c d - I C 17 TE 0.983 1.145 0.906 "C-l a t a temperature of 4O C. The thermal defect of
Far West 1.0 cm3-IC 17A TE 0.984 1.145 0.906 A-150 plastic has been measured to be + 4.0% =
Far West 0.1 cm3-IC 18 TE 0.991 1.145 0.906 1.5% (McDonald and Goodman, 1982) where the
WJe = 33.77 J/C for ambient air filling is recommended for all positive sign indicates that a fraction of the depos-
chambers. ited energy is lost to non-thermal processes. For
w,le = 34.8 J/C for ambient air filling is recommended for all water, a value of 0% f- 1.0% (Schulz et al., 1992) is
chambers. recommended and for graphite, 0% is assumed on
w,lW, = 1.031 for ambient (humidified) air-for dry air see
section 5.4.3.2.
the basis of fundamental considerations.
Khum = 0.997 for ambient air filling is recommended for all
chambers.
= 1.134 is the recommended value for the restricted 7.6 Summary of Recommendations
stopping power for water to air for W o photons.
This Section summarizes the recommendations on
the determination of proton absorbed dose which
were made in this Report:
correction (usually very small) for ion recombination 1. A standard A-150 tissue-equivalent or graphite
effects in the calibration beam is represented by Aion, thimble ionization chamber having a standard
which can be provided by the standards laboratory 60Co calibration factor is the recommended
along with the calibration factor. reference dosimeter for clinical proton dosirn-
For s , &for protons, values given in ICRU Report etry. The chamber should be open and filled
49 (ICRU, 1993) should be used (see Table 7.1). For with ambient air. Chambers of greater than 0.5
( w & , l e above about 1MeV, a value of 34.8 + 0.7 JIC cm3 volume can be used in large diameter
is recommended as discussed in Section 5.4 of this beams, but smaller chambers must be used in
Report. While the earlier ICRU recommended value small beams. Particularly for TE chambers,
of 35.2 ( 5 4%) JIC was based on low energy measure- additional measurements with TE-methane gas
ments (ICRU, 1979), the recommendation of this filling in the chamber can be a useful confirma-
Report includes higher energy measurements in tion of the chamber operation.
nitrogen, TE methane and air (Petti et al., 1986; 2. The recommended numerical values of the re-
Hiraoka et al., 1988; Denis et al., 1990; Delacroix et quired quantities are those discussed in Section
al., 1994; Seuntjens et al., 1994; Siebers et al., 1995). 7.5 and available in reports and papers cited in
Despite the scatter in the data, summarized in the text. Tables 7.1 and 7.2 summarize these
Figures 5.8 and 5.9, these experimental results recommendations. In a polychromatic proton
confirm the relative insensitivity of w to particle beam, the stopping powers needed should be
energy when the speed is substantially above orbital evaluated a t the energy corresponding to t h e
electron speeds. The absolute value of wle enters residual csda range of protons from the point of
directly into ionization chamber dose determina- measurement to the point beyond the Bragg
tions and is estimated to introduce an uncertainty of peak where the dose falls to 10% of its m a u -
no more than i 2%. mum value, as defined in Section 7.5.
Both s,&,and (w&le are dependent on energy. 3. When possible, a water, graphite or A-150
However, as can be seen in Figure 5.5 and Table 7.1, calorimeter should be used to confirm the pro-
the variation in s , , ~ ,from 6 MeV to 172 MeV (csda ton calibration factor of the reference chamber.
water range from 0.5 mm to 20 cm) is only about Thermal defect corrections, as discussed in
1.2%. Therefore, it is recommended that the residual Section 7.3, should be used.
csda range of the beam in water be used for selecting 4. A fluence-based dosimetry technique, such as a
the effective energy for s , & , As a practical definition Faraday cup can be a useful way to indepen-
of residual csda range, the distance in water between dently verify the calibration of the reference
the point of measurement and the depth at which the chamber if the energy and effective area of the
dose falls to 10%of its maximum value, can be used. beam can be accurately determined or calcu-
For a modulated beam, this will slightly underesti- lated. Such a technique can be useful when a
mate the value of (Slp), but by no more than about calorimeter is not available, although it should
 be recognized that the uncertainties of Auence-
based dosimetry tend to be large.
5. The proton absorbed dose, as determined by the
techniques described, should be specified in
water. Apart from range corrections, the phan-
tom material is not of critical importance, al-
though it can affect the probability of nonelastic
nuclear interacti~ns.Therefore, it is recom-
mended that water or other tissue-like phan-
toms be used.
6 . Measurements should be made in a phantom a t
a point where the dose is rather uniform. The
beam size should be large enough so that a n
increase in field size does not significantly
change the detector response. Measurements
should be made a t the center of the SOBP
(Spread-Out-Bragg-Peak) in the clinical beams.
The phantoms should have transverse dimen-
sions significantly larger than the cross-section
of the beam.
7. Since agreement on beam calibration among
proton facilities is of the utmost importance
for comparison of clinical results, efforts
should be made 6 compare proton doses in
various institutions. This can be done through
direct intercomparisons with ionization cham-
bers, or indirectly through the use of traveling
calorimeters, Faraday cups or even mailed inte-
grating dosimeters. When possible, dosimetric
intercomparisons should be combined with mi-
crodosimetric and radiobiological intercompari-
sons.
 Appendix A
Quality Assurance
A program of quality assurance of the physical
aspects of proton beam treatment is essential. The
quality assurance program should have formal ap-
proval and documentation should be available for
periodic review by outside agencies and consultants.
Following commissioning of a new treatment sys-
tem, a set of periodic quality assurance procedures
must be established. The frequency of check proce-
dures should be high at the initiation of patient
treatment and reduced frequencies considered only
after significant experience is gained. All alterations
of check procedures and frequencies should be sub-
ject to formal approval.
Each proton treatment facility will have differ-
ent quality assurance requirements. The items
listed below are offered as guidelines and sugges-
tions and will not describe all tests that must be
made a t any particular facility. These tests are
clinical physics procedures and do not include engi-
neering and maintenance tests or checks made to
assure proper alignment of individual patients for
daily treatment. Some items in the list will not
be relevant to all types of accelerator and beam
delivery systems. The frequency of checks will vary
depending on the stability of individual treatment
systems.
Shown below is a list of items included routinely in
quality assurance checks at the proton facility at the
Loma Linda University Medical Center. This list is
only an example.
New treatment portals
* Dose monitor calibration
* Consistency of planned treatment range with
beam energy and absorber thicknesses
* Coincidence of the physical aperture shape with
the planned shape
* Qualitative consistency of the range com-
pensator shape with treatment plan isothick-
ness lines
* Comparison of range compensator physical thick-
nesses measured at selected points with planned
thickness
* Agreement of dose monitor calibration with
predictions from a computational model
Daily checks
* Patient alignment aids including lasers and
radiographic systems
* Beam range for each acceleration energy
* Dose monitor calibration for specific sets of
standard conditions
* Individual patient treatment calibration and
range check
* Uniformity of range modulation
* Backup dose monitor function
* Changes in beam monitor target and tolerance
parameters
Weekly checks
* Dose delivered for a random selection of sample
patient treatments
* Treatment field lateral uniformity
* Room interlock checks
Biannual checks
* Intercomparison of beam calibration detectors
Yearly checks
* Patient alignment systems
* Gantry isocenter location and uncertainty
* Patient support system motions and readouts
* Depth dose distribution for monoenergetic and
modulated beams
* Lateral uniformity and penumbra measure-
ments for a range of conditions
* Dependence of dose monitor calibration on treat-
ment parameters
* Base line data for daily quality assurance checks
* Calibration intercomparison with other proton
facilities
 Appendix B
Dosimetry Worksheet
This worksheet can be used for determining the absorbed dose in water for a clinical proton beam, using an
ionization chamber calibrated in a 6Wobeam.
NAME: INSTITUTION:
DATE:
I. Ionization chamber data
(a) Chamber model and serial number:
(b) Total wall + cap thickness: g
(c) AWau (Table 7.2)
(d) s w d l , a i r (Table 7.2)
(el ( ~ J p ) , r , w a ~ ~ (Table 7.2)
(f) Aim(calibration laboratory)
(g) NK(air-kerma calibration factor) J kg-l reading-I
(h) N, (exposure calibration factor) R reading-l
(i) ND,,,(absorbed dose to water factor) Gy reading-I
Cj) Khum (humidity correction factor)
[Note-if calibrating laboratory returns only N,, then calculate NK = Nx(2.58 X 10-4)(WJe)/(l-g)where g =
0.003 and (WJe) = 33.77 JIC (see text)]

(k)ND,a, (chamber factor) Gy reading-'

11. Proton beam data
(a) Initial proton energy: MeV
(b) Collimator field size: cm2
(c) Width of SOBP: cm
(d) Range (R)in water to 10% dose pt. cm
(e) Depth (D) of measurement in water cm
(f) Residual range (R-D) cm
(g) Stopping Power ratio (s~,,~),
a t this energy (from Table 7-1)
(h) wIJWC (from Table 7.2)
(i) Cp = (s,,,)~ E(~ar)dWalr)cl
6 ) k, = Cd(sW,,,), (from Table 7.2)
111. Absorbed dose in the proton beam
(a) Temperature T: OC
(b) Pressure P: kPa
(c) p ~ , p= [(T+273.15)/(Tref+273.15)1 X (Prep)
where Tref and Prefcorrespond to the temperature and pressure conditions for the 'j°Co reference
calibration
(d) p,,, (ion collection efficiency in proton beam)
Reading Monitor Units Readingmu
(M) (MU) M/MU

(e) pj (other correction factors) humidity

polarity
other
(f) Ptot = PT,PX Pion X IT,(Pj)-
(product of all correction factors)
I Mean (MIhW
( g ) McoJMU= Mean (M/MU) X pmt
For kerma-based calibrations:
(h) D n U = (McoJMU)N ~ , a i Cp:
r Gy MU-'
or
For absorbed dose to water calibrations:
(i) DwNU = (McoJMU)N D , k~p, ~ Gy MU-'
 References

AAPM (1983). "A protocol for the determination of cyclotron system for proton therapy," Nucl. Irzst.
absorbed dose from high-energy photon and elec- Meth. Phys. Res. B B56/57,1201.
tron beams," Med. Phys. 10,741. BENNETT,G. W. and ARCHAMBEAU, J. 0. (1977).
AAPM (1986). American Association of Physicists in "Experimental tests of proton beam localization,"
Medicine. Protocol for heavy charged-particle Med. Phys. 4,115.
therapy beam dosimetry, AAPM Report No. 16 BETHE,H. A. (1953). "Moliere's theory of multiple
(Americaa Association of Physicists in Medicine, scattering," Phys. Rev. 89,1256.
American Institute of Physics, New York). BICHSEL, H. (1968). "Charged-particle interactions,"
ABROSIMOV, N. K., VOROBEVL, A. A., ZHERBIN, E. A. and p. 157-228 in Radiation Dosimetry, vol. I, F. H.
KONNOV,B. A. (1985). "Proton therapy a t the Attix, W. C. Roesch and E. Tochilin Eds. (Academic
cyclotron in Gatchina, USSR," Proc. Acad. Sci. Press, New York).
USSR 5,84. BICHSEL,H. (1992). "Stopping power and ranges of
AMALDI,U., ARDUINI,G., CAMBRIA, R., CAMPI,D., fast ions in heavy elements," Phys Rev A 46,5761.
CANZI,C., GERARDI, F., GRAMATICA, R., LEONE,R., BICHSEL,H. (1996). "Calculated Bragg curves for
MANFREDI, G., NINIS,M., PETRUCCI, G., ROSSI,S., ionization chambers of different shapes," Med.
SANGALETTI, L., SILARI, M., ' I ~ s IG.,
, VECCHI, L. and Phys. 22,1721-1726.
WEISS,M. (1994). "Ahospital-based hadrontherapy BICHSEL,H. and HIRAOKA, T. (1989). "Energy spectra
facility for Italy," p. 329-339 in Hadrontherapy in and depth dose curves for 70 MeV protons," Int. J.
Oncology, U. Amaldi and B. Larsson Eds. (Elsevier, Quantum Chem. Symp. 23,565.
Amsterdam). BICHSEL,H., HANSON,K. M. and SCHILLACI,M. E.
ANDERSEN, H. H. and ZIEGLER, J. F. (1977). Hydrogen (1982). "Lateral displacement in small angle mul-
Stopping Powers and Ranges I n Elements (Pergam- tiple scattering," Phys Med Biol27,959.
mon, New York). BIPM BUREAU INTERNATIONAL DES POIDSET MESURES
ATTM, F. H. (1986). Introduction to Radiological (1985). Constantes physiques pour les etalons de
Physics a n d Radiation Dosimetry (John Wiley, mesure des rayonnements (Com. Cons. Etalons.
New York). Mes. Ray. Ionisants, 92312 Cedex).
AUSTIN-SEYMOUR, M., MUNZENRIDER, J. E., GOITEIN, B u m , H. and CORAY, A. (1990). "A horizontal
M., GENTRY,R., GRAGOUDAS, E., KOEHLER, A. M., proton beam line for the development of a scan-
MCNULTY, P., OSBORNE, E., RYUGO,D. K., SEDDON, ning technique,"Radiotherapy a n d Oncology 17,17.
J., URIE,M., VERHEY,L. and SUIT,H. D. (1985). B L A ~H., ,CORAY, A., PEDRONI, E. and GREINER,
"Progress in low-LET heavy particle therapy: R. (1990). "Spot scanning for 250 MeV protons,"
intracranial and paracranial tumors and uveal Strahlentherapie a n d Onkologie 166,45.
melanomas,"Radiat. Res. 104, S219-S226. BLOSSER,H., JOHNSON, D., LAWTON, D., RoNNIXGEN,
AUSTIN-SEWOUR, M., MUNZENRIDER, J. E., GOITEIN, R., BURLEIGH, R. and GOTTSCHALK, B. (1991). Na-
M., VERHEY, L. and URIE,M. (1989). "Fractionated tional Superconducting Cyclotron Laboratory.
proton radiation therapy of chordoma and low- Michigan State University. Preliminary design
grade chondrosarcoma of the base of the skull," J. study exploring building features required for a
Neurosur. 70,13. proton therapy facility for Ontario Cancer Insti-
AUSTIN-SEYMOUR, M., MUNZENRIDER, J., LINGGOOD, R., tute, MSUCL 760a (National Superconducting Cy-
GOITEIN,M., VERHEY, L., URIE, M., GENTRY,R., clotron Laboratory, Michigan State University,
BIRNBAUM, S., ROUTOLO, D. and CROWELL, C. (1990). East Lansing, MI (USA)).
"Fradionated proton radiation therapy of cranial and BOAG,J. W. (1966). "Ionization Chambers," p. 1 in
intracranial t u m o w , " h . J.Clin. Oncol. 13,3274330. Radiation Dosimetry, vol. 11, F. H. Attix and. ' 1 C.
BAKKER,C. and SEGRE,E. (1951). "Stopping power Roesch Eds. (Academic Press, New York).
and energy loss for ion pair production for 340- BOONE,M. L. M., LAWRENCE, J. H., C O N N OIT ~.G.,
MeV protons," Phys. Rev. 81,489. MORGADO, R., HICKS,J . A. and BROWN, R. C. (1977).
BACHER, R., BLATTMANN, H., BOEHRINGER, T., CORAY, "Introduction to the use of protons and heavy ions
A., EGGER,E., PEDRONI, E., PHILLIPS, M. and SCHEIB, in radiotherapy: historical perspective," Int. J.
S. (1989). "Development and first results of dis- Radiat. Onc. Biol. Phys. 3,65.
crete dynamic spot scanning with protons," p. 9 in BORING, J. W., STROHL, G. E. and WOODS, F. R. (1965).
Heavy Particle Therapy Workshop (PTCOG, "Total ionization in nitrogen by heavy ions of
EORTC, ECNEU), PSI, (Report No. 69, Paul Sch- energies 25 to 50 keV," Phys. Rev. A 140,1065.
errer Institute, CH-5232, Villigen). BORTNER, T. E. and HURST,G . S. (1953). "Energy per
BEECKMAN, W., JONGEN, Y., LAISNE,A. and LANNOYE, ion pair for 5 MeV alpha particles in helium," Phys.
G. (1990). "Development of a low-cost compact Rev. 90,160.
BOUTILLON, M. (1987). CCEMRI, Offilib. Values of g DAVIS,L., HOEL,D., FOX,J. and LOPEZ,A. (1990).
for photon energies, 85-18 (CCEMRI, Offilib, Paris). "International trends in cancer mortality in France,
BOUTILLON, M. and PERROCHE-ROUX, A. M. (1987). West Germany, Italy, Japan, England and Wales.
"Re-evaluation of the W value for electrons in dry and the USA," The Lancet 336,474.
air," Phys. Med. Biol. 32,213. DELACROM, S., BRIDIER, A., W,A., DAURES, J..
BOYD,T. J., C R ~ ~ ~ AK.LR., L ,WM, R. W., HANSBOR- OSTROWSKY, A., KACPEREK, A., WNCKIER,S., BRAS-
OUGH, L. D., HOEBERLING, R. F., JAMESON,R. A., SARD, N., HABmnd, J. L. and NAURAYE, C. (1994).
KNAPP, E. A., MUELLER,D. W., POITER, J. M., "Proton dosimetry comparison involving ionom-
STOKES,R. H., STOVALL, J . E., STURGESS, R. G., etry and calorimetry," in NPL Calorimetry Work-
SWENSON, D. A., TALLERICO, P. J., WANGLER, T. P. shop, NPL, Teddington, UK,
and WILKERSON, L. C. (1982). "The PIGMI linear DENIS,J. M., SLYPEN, I., TILQUIN, I. and MEULDERS.
accelerator technology,"p. 3 in Pion and Heavy Ion J.-P. (1990). "Average ionization energy, W,for 6 5
Radiotherapy: Pre-Clinical Studies, E. L. D. SKARS- MeV protons in nitrogen," page 515 in 2nd Euro-
GARD Ed. (Elsevier, Amsterdam). pean Particle Accelerator Conferece, Medical Satel-
CASE,J. A., VONGOELER, G., RABIN,M.S.Z., and WAG lite Meeting, Nice, June 14-16,1990 P. Chauvel, A
NER, M. (1994). Proton Therapy Cooperative Group Wambersie and P. Mandrillon Eds. (Editions Fron-
Meeting. Effects ofDetector Geometry and Sensitiv- tiers, Gif-sur-Yvette, France).
ity on the Shape of the Bragg Peak, (Proton Therapy DEVITA,V. T. (1983). "Progress in cancer manage-
Cooperative Group Meeting, Cambridge, MA). ment," Cancer 51,2401.
CASTRO,J. R., CHEN, G . T, Y. and BLAKELY, E. A. DEVITA,V. T. and KORN,D. (1986). "Progress against
(1985). "Current considerations in heavy charged- cancer," New Engl. J.Med. 315,964.
particle radiotherapy: A clinical research trial of DEVITA,V. T., HELLMAN,S. and ROSENBERG, S. A
the University of California Lawrence Berkeley (1989). Cancer: Principles and practice of oncology
Laboratory, Northern California Oncology Group (Lippincott, Philadelphia).
and Radiation Therapy Oncology Group," Rad. DOLL,R. (1990). "Are we winning the fight against
Res. 104, S263. cancer? An epidemiological assessment, EACR-
CAWS, J., OSTROWSKY, A., STEINSCHADEN, K, him- Muehlboch Memorial Lecture," Eur. J. Cancer 26,
CAUX, M., CANCE,M., SIMOEN, J. P., SABAT~IER, R. 500.
and BRETEAU,N. (1984). "Direct calibration of DOMEN,S. R. (1980). "Absorbed dose water calorim-
ionization chambers with a TE calorimeter at the eter," Med. Phys. 7,157.
Orleans cyclotron neutron facility," Strahlenthera- DOMEN,S. R. (1986). "Advances in calorimetry for
pie 160,127. radiation dosimetry," p. 245 in The Dosimetry of
CHEMTOB,M., PARMENTIER, N. and NGUYEN,V. D. Ionizing Radiation, vol. 11, K R. Kase, B. E.
(1978). "Some experimental results on W values Bjarngard and F. H. Attix Eds. (Academic Press,
for heavy particles," Phys. Med. Biol. 23,1197. New York).
CHU,W., McEvoy,M., NYMAN, M., RENNER, T., Gems, DOMEN,S. R. (1994). "A sealed water calorimeter for
B., SINGH, R. P. and STRADTNER, R. (1985). 'Wobbler measuring absorbed dose,"J.Res. Natl. Inst. Stand.
dosimetry for biomedical experiments a t the BEV- Technol. 99,121.
ALAC,"IEEE Duns. on Nucl. Sci. NS-32,3324. DOMEN, S. R. and LAMPERTI, P. J. (1974). "A heat-loss
CHU,W. T., RENNER, T. R. and LUDEWIGT, B. A. (1989). compensated calorimeter: Theory, design, and per-
"Dynamic beam delivery for three-dimensional formance,"J. Res. Natl. Bur. Stand. 7 8 4 595.
conformal therapy," p. 295 in EULIMA Workshop DUDLEY, R. A. (1966). "Dosimetry with photographic
on the Potential Value of Light Ion Beam Therapy, emulsions," p. 325 in Radiation Dosimetry, vol. 11.
Nice, France, P. Chauvel and A. Wambersie Eds., F. H. Attix, W. C. Roesch and E. Tochilin Eds.
Publication No. EUR 12165 EN (Commission of (Academic Press, New York).
the European Communities, Brussels-Luxembourg). EVANS,R. D. (1955). The Atomic Nucleus (McGraw-
COLE,F., LIVDAHL, P.V.,MILLS,F. and TENG,L. (1987). Hill, New York).
"Design and application of a proton therapy accel- FLEMING, D. M. and GLASS,W. A. (1969). "Endother-
erator," p. 1985 in Proceedings of the1987 IEEE mic processes in tissue-equivalent plastic," Ra -
particle accelerator conference, E. R. Lindstrom diat. Res. 37,316.
and L. S. a y l o r Eds. (IEEE Press, Piscataway, NJ). FUKLTMOTO,S., ENDO,K , MUTO,K., AKISADA,M.,
CONSTABLE, I. J. and KOEHLER, A. M. (1974). "Experi- KITAGAWA, T., INADA,T., TSUJII,H., MARUHASHI, k,
mental ocular irradiation with accelerated pro- HAYAKAWA, Y., TAKADA, Y. and TADA,J. (1989 1.
ton," Invest. Ophthal. 13,280. "Tsukuba Medical Proton Synchrotron," p. 70 in
COUTRAKON, G., MILLER,D., KROSS,B. J., ANDERSON, Heavy Particle Therapy Workshop (PTCOG.
D. F., DELUCA, P. M. and SIEBERS,J. (1991). "A EORTC, ECNEU), PSI, (Report No. 69, Paul Sch-
beam intensity monitor for the Loma Linda cancer emer Institute, CH-5232,Villigen).
therapy proton accelerator," Med. Phys. 18,817. GASTORF,R., HUMPHRIES, L. and ROZENFELD, M.
 (1986). "Cylindrical chamber dimensions and the LARSSON, B., LIUA,A., PELETTIERI, L., REISTAD,D.,
corresponding values of A,,,, and N,,,1 (Nx Aion)," RIKNER, G., RUSSELL, K, BLOM,M. and MEDIN,J.
Med. Phys. 13,751. (1991). Paul Scherrer Institute, CH-5232, Villigen.
GOITEIN,M. and ABRAMS, M. (1983). "Multi-dimen- The proton therapy project of the Svedberg Labora-
sional treatment planning: I. Delineation of anatomy," tory in Uppsala, 111 (Paul Scherrer Institute,
Int. J.Radiat. Oncol. Bwl. Phys. 9,777-787. CH-5232, Villigen, Switzerland).
GOITEIN, M. and MILLER, T. (1983). "Planning proton GRUSELL, E., ISACSSON, U., MONTELIUS, A. and MEDIN,
therapy of the eye," Med. Phys. 10,275. J. (1995). "Faraday cup dosimetry in a proton
GOITEIN,M., ABRAMS,M., ROWELL, D., POLLARI, H. therapy beam without collimation," Phys. Med.
and WILES,J. (1983). "Multi-dimensional treat- Biol. 40,1831.
ment planning: 11. Beam's eye-view, back projec- GUEULETTE, J., GREGOIRE, V., OCTAVE-PRIGNOT, M.
tion, and projection through CT sections," Int. J. and WAMBERSIE, A. (1996). "RBE measurements in
Radiat. Oncol. Biol. Phys. 9,789-797. the 85 MeV proton beam produced a t the cyclotron
GOODMAN, L. J. (1969). "Amodified tissue equivalent CYCLONE of Louvain-la-neuve (Belgium)," Rad.
liquid," Health Phys. 16,763. Res. 145,70-74.
GOODMAN, L. J. and COYNE,J . J. (1980). "W, for GULBRANDEN, T. and MADSEN, C. B. (1962). "Radia-
Methane-Based Tissue-Equivalent Gas," p. 255 in tion dosimetry by means of semiconductors,"Acta.
Proceedings of the Workshop on Ion Chambers for Radiol. 58,226.
Neutron Dosimetry, J. J. Broerse Ed. (Harwood HABERER, T., BECHER,W., SCHARDT, D. and KRAET,G.
Academic Publishers, Rijswijk). (1993). "Magnetic scanning system for heavy ion
GOODMAN, G. B., DIXON,P., LAM,G. K. Y., HARRISON, therapy," Nucl. Instr. Meth. in Phys a s . (A) 330,
R., KORNELSEN, R. O., LUDGATE, C. M. and FLORES, 296-305.
A. D. (1985). "Preparatory clinical studies of pi- HALL,E. J., KELLERER, A. M., ROSSI,H. H. and LAM,Y.
mesons a t TRIUMF," Rad. Res. 104, S279-5284. P. (1978). "The relative biological effectiveness of
GORE,J. C., KANG,Y. S. and SCHULZ,R. J. (1984). 160 MeV protons. 11. Biological data and their
"Measurement of radiation dose distributions by interpretation in terms of microdosimetry," Int. J.
nuclear magnetic resonance (NMR)imaging," Phys. Radiat. Oncol. Bwl. Phys. 4,1009-1014.
Med. Biol. 29,1189. HAMM,R. W., CRANDALL, K. R. and POTTER, J. M.
GRAFFMAN, S. and JUNG,B. (1970). "Clinical trials in (1991). "Preliminary design of a dedicated proton
radiotherapy and the merits of high energy pro- therapy linac," p. 2583 in IEEE Particle Accelera-
tons,"Acta. Radiol. Ther. Phys. Biol. 9 , l . tor Conference: Accelerator Science and Technol-
GRAFFMAN, S., JUNG,B., NOHRMAN, B. A. and BERG- ogy, Lawrence Berkeley Laboratory, Berkeley, CA
STROM,R. (1967). "Supplementary treatment of (USA), L. Lizama and J. Chew Eds. (IEEE Press,
nasopharyngeal tumours with high energy pro- Piscataway, NJ).
tons,"Acta Radiol. Ther. Phys. Biol. 6,361. HANSEN, J. W. and OLSEN,K. J. (1985). "Theoretical
GRAFFMAN, S., HAYMAKER, W., HUGOSSOHN, R. and and experimental radiation effectiveness of the
JUNG, B. (1975). "High-energy protons in the post- free radical dosimeter alanine to irradiation with
operative treatment of malignant gliomas," Acta heavy charged particles," Rad. Res. 104,15.
Radiol. Ther. Phys. Bwl. 14,443. HANSEN,J. W. and OLSEN,K. J. (1989). "Predicting
GRAFFMAN, S., BWME, A. and LARSSON, B. (1985). decay in free-radical concentration in L-alpha-
"Proton radiotherapy with the Uppsala cyclotron. Alanine following high-LET radiation exposure,"
Experience and plans,"Strahlentherapie 161,764. Appl. Radiat. Isot. 40,935.
GRAGOUDAS, E. S., GOITEIN,M., KOEHLER,A. M., HAYAKAWA, Y. and SCHECHTMAN, H. (1988). "Com-
VERHEY, L., ' ~ P P E RJ.,
, SUIT,H. D., BROCKHURST, R. ments on the value of the average energy ex-
and CONSTABLE, I. J. (1977). "Proton irradiation of pended per ion pair formed in air for a proton beam
small choroidal malignant melanomas," Amer. J. recommended by the American Association of
Ophthal. 83,665473. Physicists in Medicine," Med. Phys. 15,778.
GRAGOUDAS, E., KOEHLER, A. M., URIE,M., GENTRY, R., HAZLE,J. D., HEFNER, L., NYERICK, C. E., WILSON,L.
SHIPLEY,W., URANO,M., DUTTENHAVER, J. and and BOYER,A. L. (1991). "Dose response character-
WAGNER,M. (1982). "Evaluation of the clinical istics of a ferrous-sulfate-doped gelatin system for
applicability of proton beams in definitive fraction- determining radiation absorbed dose distributions
ated radiation therapy," Int. J. Radiat. Onc. Bwl. by magnetic resonance imaging (Fe MRI)," Phys.
Phys. 8,2199. Med. Biol. 36,1117.
GRUSELL, E. and RIKNER, G. (1986). "Evaluation of HIRAOKA, T., KAWASHIMA, K. and HOSHINO, K. (1988).
temperature effects in p-type silicon detectors," "Determination of differential w-value for proton.
Phys. Med. Biol. 31,527. deuteron, 3He, neutron and 60Co gamma-ray beams
GRUSELL, E., MONTELIUS, A., BLOMQUIST, E., BRAHME, in several gases (abstract),"Phys. Med. Bwl. 33,131.
A., CARLSSON, A. C., JAKOBSSON, P., JUNG, B., HOHLFIELD, K. (1988). "The standard DIN 6800:
 Procedures for absorbed dose determination in JESSE,W. P. and SADAUSKIS, J. (1952)."Alpha-particle
radiology by the ionization method (IAEA-SM-2981 ionization in mixtures of the noble gases," Phys.
31); p. 13-22 in Dosimetry i n Radiotherapy, Vi- Rev. 88,417.
enna, 1987, (IAEA, Vienna). JESSE,W. P. and SADAUSKIS, J. (1953)."Alpha-particle
IAEA (1987). International Atomic Energy Agency. ionization in pure gases and the average energy to
Absorbed Dose Determination i n Photon and Elec- make a n ion pair," Phys. Rev. 90,1120.
tron Beams: A n International Code of Practice, JONGEN, Y. (1991). Paul Scherrer Institute. Status
Technical Series Report No. 277 (International report of IBA's proton therapy program, 111 (Paul
Atomic Energy Agency, Vienna). Schemer Institute, Villigen, Switzerland).
ICRP (1959). International Commission on Radiologi- KACPEREK, A. and BONNETT, D. E. (1989). "Develop-
cal Protection. Report of Committee I1 on Permis- ment of a Faraday cup for proton beam dosimetry
sible Dose for Internal Radiation, ICRP Publica- a t the MRC Cyclotron Unit a t Clatterbridge Hospi-
tion 2 (International Commission on Radiological tal," p. 53 in Proceedings of the International Heavy
Protection, Pergamon Press, Oxford, England). Particle Therapy Workshop, (Report 69, Paul Schemer
ICRU (1964). International Commission on Radia- Institute, CH-5232, Vigen,Switzerland).
tion Units and Measurements. Physical aspects of KANAI,T., KAWACHI, K., KUMAMOTO, Y., OGAWA, H.,
radiation, ICRU Report lob (International Com- YAMADA, T., MATSUZAWA, H. and INADA,T. (1980).
mission on Radiation Units and Measurements, "Spot scanning system for proton radiotherapy,"
Bethesda, MD). Med. Phys. 7,365.
ICRU (1977). International Commission on Radia- KANAI,T., KAWACHI, K., MATSUZAWA, H. and INADA, T.
tion Units and Measurements. Neutron Dosimetry (1983a). "Broad beam three-dimensional irradia-
for Biology and Medicine, ICRU Report 26 (Inter- tion for proton radiotherapy," Med. Phys. 10,344.
national Commission on Radiation Units and Mea- KANAI,T., KAWACHI, K., MATSUZAWA, H. and INADA, T.
surements, Bethesda, MD). (1983b). "Three-dimensional beam scanning for
ICRU (1979). International Commission on Radia- proton therapy,"iVucl. Inst. Meth. 214,491.
tion Units and Measurements. Average Energy KAPCHINSKII, I. M. and TEPLYAKOV, V. A. (1970).
Required to Produce a n Ion Pair, ICRU Report 31 "Linear ion accelerator with spatially homogeneous
(International Commission on Radiation Units strong focussing," Prib. Tekh. Eskp. 19,19-22.
and Measurements, Bethesda, MD). KARZMARK, C. J. (1964). "Secondary emission moni-
ICRU (1980). International Commission on Radia- tor as a linear accelerator electron beam dose
tion Units and Measurements. Radiation Quanti- monitor," Rev. Sci. Instr. 35,1646-1652.
ties and Units, ICRU Report 33 (International KELLERER, A. M., HAHN,K. and Rossr, H. H. (1992).
Commission on Radiation Units and Measure- "Intermediate dosimetric quantities," Radiat. Res.
ments, Bethesda, MD). 130915-25.
ICRU (1984a). International Commission on Radia- KHOROSHKOV, V. S., GOLDIN, L. L., ONOSOVSKY, K. K.
tion Units and Measurements. Radiation Dosim- and KLENOV, G. 1. (1991). "Soviet project of proton
etry: Electron beams with Energies between 1 and therapy facility with several treatment rooms," p.
50 MeT/: ICRU Report 35 (International Commis- 37 in Proceedings of the Proton Radiotherapy
sion on Radiation Units and Measurements, Workshop, PSI, (Report 111, Paul Scherrer Insti-
Bethesda, MD). tute, CH-5232, Villigen, Switzerland).
ICRU (1984b). International Commission on Radia- KJELLBERG, R. N., SWEET, W. H., PRESTON, W. M. and
tion Units and Measurements. Stopping Powers KOEHLER, A. M. (1962). "The Bragg peak of a
for Electrons and Positrons, ICRU Report 37 (Inter- proton beam in intracranial therapy of tumors,"
national Commission on Radiation Units and Mea- Duns. Amer. Neurol. Assoc. 87,216.
surements, Bethesda, MD). KJELLBERG, R. N., HANAMURA, T., DAVIS,K. R., LYONS.
ICRU (1989). International Commission on Radia- S. L. a n d h ~ ~R.sD. , (1983). "Bragg-peak proton-
tion Units and Measurements. Tzssue Substitutes beam therapy for arteriovenous malformations of
i n Radiation Dosimetry and Measurement, ICRU the brain,"New Eng1.J. Med. 309,269.
Report 44 (International Commission on Radia- KLEVENHAGEN, S. C. (1977). "The non-linearity of the
tion Units and Measurements, Bethesda, MD). temperature response of silicon p-n junction radia-
ICRU (1993). International Commission on Radia- tion detectors operated in the dc mode," Phys. Med.
tion Units and Measurements. Stopping Powers Biol. 22,353.
for Protons and Alpha Particles, ICRU Report 49 KLIAUGA, P. J., COLVETT, R. D., LAM,Y. L. and Ross,
(International Commission on Radiation Units H. H. (1978). "The relative biological effectiveness
and Measurements, Bethesda, MD). of 160 MeV protons I. Microdosimetry," Int. J .
JANNI, J. F. (1982). "Proton range-energy tables, 1 Radiat. Oncol. Biol. Phys. 4,1001-1008.
keV-10 GeV, Part I," Atomic Data and Nuclear KLIGERMAN, M. M., VONESSEN, C. F., KHAN, M. K..
Data Tables 27,147-339. SMITH, A. R., STERNHAGEN, C. J. and SALA,J. M.
 (1979). "Experience with pion radiotherapy," Can- energy per ion pair for hydrogen and oxygen ions in
cer 43,1043-51. a tissue-equivalent gas," Radiut. Res. 55,l.
KLIMAN,B., KJELLBERG, R. N., SWISHER, B. and BUT- LEVIN,C. P., GONIN,R. and WYNCHANK, S. (1988).
LER, W. (1984). "Proton beam therapy of acro- "Computed limitations of spot scanning for therapeu-
megaly: a 20-year experience," in Secretory 5%- tic proton beams,"Int. J.Biomed. Comput. 23,33.
mors of the Pituitary Gland (Progress in Endocrine LIN, S., BLAWMANN, H., BOEHRINGER, T., CORAY, A.,
Research and Therapy), vol. I, P. M. Black Ed. KOHLI,M. and PEDRONI, E. (1994). "Preparatory
(Raven Press, New York). work for the protontherapy beam dosimetry sys-
KNOLL,G. I? (1989). Radiation Detection and Mea- tem at PSI," p. 481 in Hadrontherapy in Oncology, U.
surement (John Wiley, New York). Amaldi and B. Larsson Eds. (Elsevier,Amsterdam).
KOEHLER, A. M. (1967). "Dosimetry of proton beams Lomy, R. A. and -R, G. H. (1958). "Ionization yield
using small silicon diodes,"Rad. Res. Supp. 7,53. of protons in nitrogen and argon," Phys. Rev. 109,826.
KOEHLER, A. M., SCHNEIDER, R. J. and SISTERSON, J. MARTIN, R. L. (1987). "A proton accelerator for medical
M. (1975). "Range modulators for protons and applications,"Nucl. Inst. Meth. Phys. B 24125,1087.
heavy ions," Nuc. Inst. Meth. 131,437. WUHASHI, A. (1977). "Characteristics of a minia-
KOEHLER, A. M., SCHNEIDER, R. J. and SISTERSON, J. ture dosimeter developed for measurement of elec-
M. (1977). "Flattening of proton dose distributions trons,"Nucl. Inst. Meth. 141,87.
for large-field radiotherapy," Med. Phys. 4,297. WYANSKI, M. J., SCHULZ,R. J., IBBOTT,G. S.,
KUHN,H. and WERBA, T. (1978). "Measurements of the GATENBY, J. C., XIE, J., HORTON, D. and GORE,J. C.
energy expenditure for the production of an ion pair in (1994). "Magnetic resonance imaging of radiation
tissue-equivalent gas for heavy particles," page 85 in dose distributions using a polymer-gel dosimeter,"
Proceedings of the Third Symposium on Neutron Phys. Med. Bhl. 39,1437.
Dosimetry in Biology and Medicine, Report No. EUR- MATSUDA,T. and IN-, K. (1981). "Computer
5848, Burger, G. and Ebert, H. G., Eds. (Commission controlled multi-leaf conformation radiotherapy,"
of the European Co~nmunities,Luxembourg). Nipp. Act. Radiol. 41,965-974.
-ON, H. (1958). "Energy loss per ion pair for MCDONALD, J. C. (1987). "Calorimetric measurement
protons in various gases," Phys. Rev. 112,1927. of the carbon kerma factor for 14.6 MeV neutrons,"
-SON, B. (1961). "Pre-therapeutic physical experi- Rad. Res. 109,28.
ments with high energy protons," Brit. J. Radiol.
MCDONALD, J. C. and GOODMAN, L. J. (1982). "Mea-
34,143.
surements of the thermal defect of A-150 plastic,"
-SON, B. (1967). "Radiologicalproperties of beams of
Phys. Med. Biol. 27,229-233.
high-energy protons,"Rudiut. Res. Suppl. 7,304.
LARSSON,B. and SARBY,B. (1987). "Equipment for
MCDONALD, J. C. and DOMEN,S. R. (1986). "A-150
plastic radiometric calorimeter for charged par-
radiation surgery using narrow 185 MeV proton
beams. Dosimetry and design," Acta Oncologica ticles and other radiations," Nucl. Instr. Meth.
26,143. Phys. Res. A 252,35.
LARSSON,B., LEKSELL, L., REXED, B., SomandER, P., MCDONALD, J. C., LAUGHLIN, J. S. and FREEMAN, R. E.
WR, W. and ANDERSSON, B. (1958). "The high- (1976). "Portable tissue equivalent calorimeter,"
energy proton beam as a neurosurgical tool," Na- Med. Phys. 3,80.
ture 182,1222. MEDIN,J., ANDREO, P., GRUSELL, E., ~ W S S O NO.,,
-SON, B., LEKSELL, L., REXED,B. and SomandEq MONTELIUS, A. and ROOS,M. (1995). "Ionisation
P. (1959). "Effect of high energy protons on the chamber dosimetry of proton beams using cylindri-
spinal cord,"Acta. Radiol. 51,52. cal and plane-parallel chambers. Nw versus Nk ion
LARSSON,B., LEKSELL, L. and REXED,B. (1963). "The chamber calibrations,"Phys.Med. Bwl. 40,1161.
use of high energy protons for cerebral surgery in MINAKOVA, Y. I. (1987). "Review of twenty years
man,"Acta. Chir. Scand. 125,l. proton therapy clinical experience in Moscow," p. 1
LARSSON, B., LIDEN,K. and SARBY, B. (1974). "Irradia- in Second International Charged Particle Therapy
tion of small structures through the intact skull," Workshop (Loma Linda University Medical Cen-
Acta. Radwl. Ther. Phys. Bwl. 13,512. ter, Loma Linda, CA, USA).
LAULAINEN, N. and BICHSEL, H. (1972). "Energy re- MINAKOVA, E. I., KIRPATOVSKALA, L. E., LIASS,F. M.,
moved by delta rays from finite volumes in the SNIGIREVA, R. I. and KRYMSKI, V. A. (1983). "Proton
passage of charged particles," Nucl. Inst. Meth. therapy of pituitary adenomas," Moditainskaia
104,531. Radiol. 28, 7-14.
LEKSELL, L., LARSSON,B., ANDERSSON, B., REXED,B., MONTELIUS, A., BLOMQUIST, E., NAESER,P., B ~ E ,
SOURANDER,P. and IMAIR, W. (1960). "Lesions in the A., CARLSSON, J., CARLSSON, A. C., GRAFFMAN,s.,
depth of the brain produced by a beam of high GRUSELL, E., HALLEN, S. and JAKUBSSON, P. (1991).
energy protons,"Acta. Radiol. 54,251. "The narrow proton beam therapy unit at the Sved-
LEONARD, B. E. and BORING, J. W. (1973). "The average berg Laboratory in Uppsala,"Acta Oncol. 30,739.
MUIR,C. S. and BOYLE, P. (1990). "The burden of
cancer in Europe," Eur. J. Cancer 26,1111.
MUNZENRIDER, J. E., AUSTIN-SEYMOUR, M., BLITZER,P.
J., GENTRY,R., GOITEIN,M., GRAGOUDAS, E. S.,
JOHNSON, K, KOEHLER, A. M., MCNULTY, P., MOUL-
TON, G., OSBORNE, E., SEDDON, J. M., SUIT, H. D.,
URIE, M., VERHEY, J. L. and WAGNER, M. (1985).
"Proton therapy a t Harvard," Strahlentherapie
161,756.
NATH, R. and S c m , R. J . (1981). "Calculated
response and wall correction factors for ionization
chambers exposed to Co-60 gamma-rays," Med.
Phys. 8,85-93.
NCRP (1981). National Council on Radiation Protec-
tion and Measurement. Dosimetry of X-Ray a n d
Gamma-Ray beams for Radiation Therapy in the
Energy Range 10 keV to 50 Mey NCRP Report No.
69 (National Council on Radiation Protection,
Bethesda, MD).
NICHIPOROV, D., KOSTJUCHENKO, V., Pma, J. M. and
BENSEN,D. L. (1995). "Investigation of applicabil-
ity of alanine and radiochromic detectors to dosirn-
etry of proton clinical beams," Applied Radiat.
Isotopes 46,1355.
NGUYEN, V. D., CHEMTOB, M., CHARY, J.,POSNY,F. and
PARMENTIER, N. (1980). "Recent experimental re-
sults on W values for heavy particles," Phys. Med.
Bwl. 25,509.
PARKS, J. E., HURST, G. S., STEWART,T. E. and
WEIDNER,H. K. (1972). "Ionization of the noble
gases by protons: Jesse effects as a function of
pressure," J. Chem. Phys. 57,5467.
PEDRONI,E., BACHER, R., B L A ~ ~ H., w , BOEH-
RINGER, T., COW, A., EGGER,E., PHILLIPS,M. and
SHEIB,S. (1989). Paul Scherrer Institute. The 200
MeV proton Therapy Project a t PSI: A Status
Report, Report No. 69 (Paul Scherrer Institute,
Villigen, Switzerland).
F'ERRET,C. (1989). Swiss Institut fur Nucleurforsch-
ung. Strahlentherapie intraokularer Melanome mit
einem Protonenstrahl, TM-21-89-01 (Swiss Insti-
t u t fur Nucleurforschung, Villigen, Switzerland).
PETTI, P. L., VERHEY,L. and WILSON,R. (1986). "A
measurement of w for 150 MeV protons in nitrogen
and argon,"Phys. Med. Biol. 31,1129.
PKILLIPS,M., PEDRONI,E., B L A ~ M A N N H.,, BOEH-
RINGER, T., CORAY, A. and S C H E S.
~ ,(1992). "Effects
of respiratory motion on dose uniformity with a
charged particle scanning method," Phys. Med.
Biol. 37,223-234.
PODGORSAK,M. B. and SCHREINER,L. J. (1992).
"Nuclear magnetic relaxation characterization of
irradiated Fricke solution," Med. Phys. 19,87.
PRAEL, R. E. and LIECHTENSTEIN, H. (1989). Los
Alamos National Laboratory. User's Guide to LCS:
The LAHET Code System, LA-UR-893014 (LaAla-
mos National Laboratory, h s Alamos, NM, USA).
-IN, M. S. W. (1987). Proton Therapy Cooperative
Group. Skinltumor dose ratio as a fumtion of
source to isocenter distance for a 250 MeV proton
beam, of the Facilities Working Group (Proton
Therapy Cooperative Group, Massachusetts Gen-
eral Hospital, Boston, MA).
RAJu, M. R. (1966). "The use of the miniature silicon
diode as a radiation dosimeter," Phys. Med. Biol.
11,371.
RAJU, M. R., LYMAN, J. T., BRUSTAD, T. and TOBIAS, C.
A. (1969). "Heavy charged-particle beams," p. 151
in Radiation Dosimetry, Vol. 111, F. H. Attix and E.
Tochilin Eds. (Academic Press, New York).
RAJU, M. R., KONNOV, B. A. and KOMAROV, E. I. (1987).
'Beport on International working group meeting on
the use of protons in radiotherapy," p. 237 in Fifth
PTCOG meeting and International Workshop on Bio-
medical Accelerators, (University of California.
Lawrence Berkeley Laboratory, Berkeley, CA, USA).
REGULLA, D. F. and DEFFNER, U. (1982). "Dosimetry
by ESR Spectroscopy of Alanine," Int. J. Appl.
Radiut. Isot. 33,1101-1114.
RENNER, T. R., C m , W. T., LUDEWIGT, B., NYMAN, M.
and STRADTNER, R. (1989). "Multisegmented ioniza-
tion chamber dosimetry system for light ion
beams," Nucl. Inst. Meth. A281,640.
RIABUKHIN, I. (1982). "Status of therapy with protons
in USSR," p. S-227 in International Workshop on
Particle Accelerators in Radiation Therapy,
RIKNER, G. (1983). Silicon diodes as detectors in
relative dosimetry of photon, electron, and proton
radiutwn fields, Ph.D. Dissertation (Uppsala Uni-
versitet, Uppsala, Sweden).
R ~ RG. ,and GRUSELL, E. (1987). "General specifi-
cations for silicon semiconductors for use in radia-
tion dosimetry,"Phys. Med. Biol. 32,1109.
ROBERTSON, J. B., WILLIAMS, J. R., SCHMIDT, R. A..
LIITLE,J. B., FLYNN, D. F. and SUIT,H. D. (1975).
"Radiobiological studies of high-energy modulated
proton beam utilizing cultured mammalian cells.'
Cancer 35,1664-1677.
ROBERTSON, J. B., EADDY, J. M., ARCHAMBEAU, J . 0..
COUTRAKON, G. B., MILLER,D. W., MOYERS, M. F..
SIEBERS,J. V., SLATER, J. M. and DICELLO, J. F.
(1994). "Variation of measured proton relative
biological effectiveness as a function of initial
proton energy," p. 706 in First International Sym-
posium on Hadrontherapy in Oncology, Amster-
dam, B. L. U. Amaldi Ed. (Elsevier Science).
ROESCH, W. C. (1958). "Dose for nonelectronic equilib-
rium conditions,"Radiat. Res. 9,399.
ROGERS, D. W. 0.(1992). "The advantages of absorbed-
dose calibration factors," Med. Phys. 19,1227-1239.
ROGERS,D. W. O., ROSS, C. K., SHORTT,K. R..
KLASSEN,N. V. and BIELAJEW,A. F. (1994). "To-
wards a dosimetry system based on absorbed d m
standards (IAEA-SM-330/9),"in Measurement Assur-
ance in Dosimetry, Vienna, 1993,(IAEA,Vienna).
R~HRIG, N. and COLVETT, R. D. (1978)."Measurement of
 W for protons, helium-4 ions, and carbon-12 ions in SISTERSON, J . (1995). Harvard Cyclotron Laboratory.
tissue equivalent gas,"Radiat. Res. 76,225. "Particlesn-Newsletter of the Proton Therapy CO-
Ross~,H. H. and FAILLA,G. (1956). "Tissue equiva- operative Group (PTCOG), (Harvard Cyclotron
lent ionization chambers," Nucleonics 14,32. Laboratory, Cambridge, MA).
SAUNDERS, W. M., CHAR,D. H., QUIVEY,J . M., CASTRO, SISTERSON, J. M., CASCIO,E., KOEHLER, A. M. and
J. R., CHEN,G. T., COLLIER, J. M., CARTIGNY, A., JOHNSON, K. N. (1991). "Proton beam therapy:
BLAKELY, E. A., LYMAN, J . T. and ZINK,S. R. (1985). Reliability of the synchrocyclotron a t the H a w a r d
"Precision, high dose radiotherapy: helium ion Cyclotron Laboratory," Phys. Med. Biol. 36,285.
treatment of uveal melanoma," Int. J. Radiat. SLATER, J. M., MILLER,D. W. and ARCHAMBEAU, J. 0 .
Oncol. Biol. Phys. 11,227-233. (1988). "Development of a hospital-based proton
SCHALLER, L., HUBER,P. and BAUMGARTNER, E. (1963). beam treatment center," Int. J. Radiat. Onc. Biol.
"Messung der Arbeit pro Ionenpaar in Stickstoff Phys. 14,761.
fur Protenen und He-Ionen im Energiegebiet unter- SLATER, J . M., ARCHAMBEAU, J. O., MILLER,D. W.,
halb 1 MeV," Helu. Phys. Acta 36,113. NOTARUS,M. I., PRESTON,W. and SLATER, J. D.
SCHULZ,R. J., ALMOND,P. R., KUTCHER,G., LOEV- (1991). "The Proton Treatment Center a t Lorna
INGER, R., NATH,R., ROGERS, D. W. O., SUNTHRALIN- Linda University Medical Center: Rationale for
GAM,N., WRIGHT,K. A. and KAHN, F. M. (1986). and description of its development," Int. J. Radiat.
"Clarification of the AAPM Task Group 21 proto- Onc. Bwl. Phys. 22,383.
col," Med. Phys. 13,755-759. SMATHERS, J. B., OWE, V. A., SMITH,A. R., ALMOND, P.
SCHULZ, R. J., Wuu, C. S. and WEINHOUS,M. S. R., A m , F. H., SPOKAS, J. J., QUAM, W. M. and
(1987). "The direct determination of dose-to-water GOODMANN, L. J. (1977). "Composition of A-150
using a water calorimeter," Med. Phys. 14,790. tissue equivalent plastic," Med. Phys. 4,74.
SCHULZ,R. J., VENKATARAN~~WAN, N. and HUQ,M. S. SMITH,A. R., ROSEN, I. I. and HOGSTRUM, K, R. (1977).
(1990). "The thermal defect of A-150 plastic and "The silicon diode as a n i n vivo dosimeter for fast
graphite for low-energy protons," Phys. Med. Biol. neutrons," Int. J. Radiat. Oncol. Biol. Phys. 2,111.
35,1563. SRDOC,D. (1970). "Experimental technique of mea-
SCHULZ, R. J., HUQ, M. S., VENKATARMANAN, N. and surement of microscopic energy distribution i n
MOTAKABBIR, K. A. (1991). "A comparison of ioniza- irradiated matter using Rossi counters," Radiat.
tion-chamber and water-calorimeter dosimetry for Res. 43,302.
high-energy x-rays," Med. Phys. 18,1229. STERNHEIMER, R. M. (1966). "Density effect for t h e
SCHULZ,R. J., VERHEY,L. J., HUQ,M. S. and VENKATA- ionization loss of charged particles," Phys. Rev.
RAMANAN,N. (1992). W a t e r calorimeter dosimetry 145,247.
for 160 MeV protons," Phys. Med. Bwl. 37,947-953. SUIT,H. D. (1982). "Potential for improving survival
SELTZER, S. M. (1993). National Institute of Stan- rates for the cancer patient by increasing the efficacy
dards and Technology. A n assessment of the role of of treatment of the primary lesion," Cancer 50,1227.
charged secondaries from nonelastic nuclear inter- SUIT, H. D. and MIRALBELL,R. (1989). "Potential
actions by therapy proton beams in water, NISTIR impact of improvements i n radiation therapy on
5221 (National Institute of Standards and Technol- quality of life and survival," Int. J . Radiat. Oncol.
ogy, Gaithersburg, MD). Biol. Phys. 16,891-895.
SEUNTJEKS,H. P., PALMANS, H., VERHAEGEN, F., DENIS, SUIT, H. D. and URIE, M. (1992). "Proton beams in
J., WNCKIER, S. and THIERENS,H. (1994). 'Water radiation therapy," Jour. Natl. Cancer Inst. 84,155.
calorimetry for clinical proton beams," p. 26 i n SUIT,H. D., GOITEIN,M., BPPER, J., KOEHLER, A. M.,
NPL Calorimetry Workshop (NPL, Teddington, UK). SCHMIDT, R. A. and SCHNEIDER, R. (1975)."Explor-
SHIPLEY,W. U., TEPPER,J. E., FROUT,JR., G. R., atory study of proton radiation therapy using
VERHEY,L. J., MENDIONDO, 0 . A., GOITEIN,M., large field techniques and fractionated dose sched-
KOEHLER,A. M. AND SUIT, H. D. (1979). "Proton ules," Cancer 35,1646-1657.
radiation as boost therapy for localized prostatic SUIT,H. D., GOITEIN,M., TEPPER, J. E. and VERHEY,L.
carcinoma," J . Am. Med. Assoc. 241,1912-1915. (1977). "Clinical experience and expectation with
SIEBERS,J. V. (1990). Shielding measurements for a protons and heavy ions," Int. J. Radiat. Oncol.
230 MeVproton beam, Ph.D. Dissertation (Univer- Biol. Phys. 3,115-125.
sity of Wisconsin-Madison). SUIT,H . D., GOITEIN, M., MUNZENRIDER, J. E., VERHEY,
SIEBERS,J . V., VATNITSKY, S. M., MILLER,D. W. and L., GRAGOUDAS, E., KOEHLER, A. M., URANO,M.,
MOYERS,M. F. (1995). "Deduction of the air w- SHIPLEY, W. U., LINGGOOD, R. M., FRIEDBERG, C. and
value in a therapeutic proton beam," Phys. Med. WAGNER, M. (1980)."Clinical experience with pro-
Biol. 40,1339. ton beam radiation therapy," J . Can. Assoc. Ra-
SILVERBERG, E., BORING,C. C. and SQUIRES,T. S. diol. 31,3539.
(1990). "Cancer statistics," C A - A Cancer Journal SUIT,H., GOITEIN, M., MUNZENRIDER, J., V E R H ~L.,,
for Clinicians 40,9. BLITZER, P., GRAGOUDAS, E., KOEHLER, A. M., URIE,
 M., GENTRY, R., SHIPLEY, W., URANO,M., DUTTEN- SUIT,H. D., MENDIONDO, O., GRADOUDAS, E. S. and
HAVER, J . and WAGNER, M. (1982a). "Evaluation of KOEHLER, A. (1984)."Relative biological effectiveness
the clinical applicability of proton beams in defini- of modulated proton beams in various murim tis-
tive fractionated radiation therapy," Int. J.Radiat. sues,"Int. J.Radiat. Oncol. Bwl. Phys. 10,5091514,
Onc. Biol. Phys. 8,2199-2205. URIE,M. and GOITEIN,M. (1989). "Variable versus
SUIT,H. D., GOITEIN, M., MUNZENRIDER, J., VERHEY, fixed modulation of proton beams for treatments
L., DAVIS,K. R., KOEHLER, A., LINGGOOD, R. and in the cranium," Med. Phys. 16,593.
OJEMANN,R. G. (1982b). "Definitive radiation URIE,M., SISTERSON, J. M., KOEHLER, A. M., GOITEIX.
therapy'for chordoma and chrondrosarcoma of the M. and ZOESMAN, J . (1986). "Proton Beam Penum-
base of skull and cervical spine," J. Neurosurgery bra:effects of separation between patient and beam
56,377-385. modifying devices,"Med. Phys. 13,734.
SUIT,H. D., GOITEIN, M., MUNZENRIDER, J., VERHEY, VATNITSKY, S., MILLER,D., SIEBERS, J. and MOYERS, M.
L., URIE, M., GRAGOUDAS, E., KOEHLER,A., (1995). "Application of solid state detectors for
GOTTSCHALK, B., SISTERSON, J., TATZUZAKI, H. and dosimetry of therapeutic proton beams,"Med. Phys.
MIRALBELL, R., (1990). "Increased efficacy of radia- 22,469.
tion therapy by use of proton beam," Strahlenthera- VATNITSKY, S., SIEBERS, J. AND MILLER, D. (1996). "kg
pie 166,40. factors for ionization chamber dosimetry in clini-
SUNAGA, H., AGEMATSU, T., TANAKA, R., YOSIDA, K. and cal proton beams," Med. Phys. 23,25-31.
KOHNO,I. (1988). "Dosimetry by cellulose triac- VERHEY, L. J. and LYMAN, J. (1992). "Some consider-
etate film dosimeter in proton-beam irradiation," ations regarding w values for heavy charged-
Radioisotopes 37,84. particle radiotherapy," Med. Phys. 19, 151.
TAUFEST, G. W. and FECHTER, H. R. (1955)."Anonsatu- VERHEY, L., KOEHLER,A. M., MCDONALD, J. C., GOIT-
rable high-energy beam monitor," Rev. Sci. Instr. EIN,M., m, I., SCHNEIDER, R. and WAGNER, M.
26,229.
(1979). "The determination of absorbed dose in a
THAMES,H. D., SCHULTHEISS, T. E., HENRY,J. H.,
proton beam for purposes of charged-paraticle
TUCKER, S. L., DUBRAY, B. M. and BROCK,W. A.
radiation therapy,"Radiat. Res. 79,34.
(1992). "Can modest escalations of dose be de-
VERHEY,L. J., GOITEIN,M., MCNULTY, P., MUNZEN-
tected as increased tumor control?,"Int. J.Radiat.
R ~ E RJ.
, E. and SUIT,H. D. (1982)."Precise position-
Biol. Oncol. Phys. 22,241.
ing of patients for radiation therapy," Int. J. Ra-
THOMAS, D. J. and BURKE,M. (1986). "Measurement
diat. Oncol. Biol. Phys. 8,289.
of W values for protons in an A-150 plasticequivalent
VONESSEN,E. F., BLATTMAN, H., CRAWFORD, J. F.,
gas and in argon,"Phys. Med. Biol. 32,1119.
TOBIAS, C. A., ROBERTS, J. E., LAWRENCE, J. H., FESSENDEN, P., PEDRONI, E., PERRET, C., S u m - .
LOW-BEER, B. V. A., ANGER, H. O., BORN,J . L., M., SHORTT, K. and WALDER, E. (1982). "The PIO-
MCCOMBS, R. and HUGGINS,C. (1955). "Irradiation TRON: initial performance, preparation and expe-
hypophysectomy and related studies using 340 MeV rience with pion therapy," Int. J. Radiat. Oncol.
deuterons,"Peaceful Uses ofAtomic Energy 10,95. Biol. Phys. 8,1499-1510.
TOBIAS,C. A., LAWRENCE, J. H., BORN,J . L., MC- VYNCKIER, S., BONNETT, D. E. and JONES,D. T. L.
COMBS,R. K., ROBERTS, J. E., ANGER,H. O., LOW- (1991). "Code of practice for clinical proton dosim-
BEER,B. V.A. and HUGGINS, C. B. (1958)."Pituitary etry," Radiotherapy a n d Oncology 20,53.
irradiation with high-energy proton beams, a pre- VYNCKIER, S., BONNETT, D. E. and JONES,D. T. L.
liminary report," Cancer Res. 18,121. (1994). "Supplement to the code of practice for
TOBIAS, C. A., LAWRENCE, J. H., LYMAN, J., BORN,J . L., clinical proton dosimetry," Radiotherapy and On-
GOTTSCHALK, A., LINFOOT, J . and MCDONALD, J. cology 32,174-179.
(1964). "Progress report on pituitary irradiation," WAGNER, M. S. (1982). "Automated range compensa-
p. 19 in Second International Symposium on The tion for proton therapy," Med. Phys. 9,749.
Response of the Nervous System to Ionizing Radiation, WAMBERSIE, A. (1995). "Nouveaux rayonnements en
University of California, Los Angeles, T. J. Haley and radiotherapie. Les premieres applications thera-
R. J. Snider Eds. (Little, Brown, Boston, MA). peutiques de neutrons et protons en Belgique.," p.
TRUMP, M. A. and PINKERTON, A. P. (1967). "Applica- 293 in 1895-1995, Hundred Years of Radiology in
tion of PN junction diodes to the measurement of Belgium, R. V. Tiggelen and J . Pringot Eds. (Askle-
dose distribution of high energy radiation," Phys. pios Editeur, Museum of Radiology, Brussels).
Med. Biol. 12,573. WAMBERSIE, A., GREGOIRE, V. and BRUCHER, J . M.
BUNEMOTO, H., MORITA, S., ISHIKAWA, T., FURUKAWA, (1992). "Potential clinical gain of proton (and
S., KAWACHI, K., KANAI, T. and O m . H. (1985). heavy ion) beams for brain tumors in children,'
"Proton therapy in Japan," Radiation Res. 104, Int. J.Radiat. Oncol. Biol. Phys. 22,275.
5235. WILSON,R. R. (1946). "Radiological use of fast pro-
URANO, M., VERHEY, L. J., GOITEIN, M., TEPPER,J. E., tons," Radiology 47,487.
 ICRU Reports
ICRU Reports are distributed by the ICRU Publications' office. Information on prices and how to order
may be obtained from:

ICRU Publications
7910 Woodmont Avenue, Suite 800 .-
Bethesda, Maryland 20814
U.S.A.
Phone (301) 657-2652
FAX (301) 907-8768

Copies of the reports may also be purchased from the following:

Mrs. Brigitte Harder Dr. Minoru Takada
Konrad-Adenauer-Stra13e 26 Japan Radioisotope Association
D-37075 Gijttingen 28-45, Honkomagome 2-chome
Germany Bunkyoku Tokyo 113, Japan
Phone (0551) 22612
Dr. Torgil Moller
Kazuya Yamashita, Ph.D. Regionala Tumiirregistrat
The Japanese Society of Radiological Technology Lasarettet
Nijyo Plaza, 88 Nishinokyo, 22 185 Lund
Kitatsuboi-cho, Sweden
Nakagyo-ku, Kyoto 604
Japan
Prof. Andre Wambersie
Unit6 de Radiobiologie et Radioprotection
UCL-Cliniques St. Luc
Avenue Hippocrate, 54.69
B-1200 Brussels, Belgium
Phone (02) 764.54.68

The currently available ICRU Reports are listed below.

ICRU
Report No. Title
lob Physical Aspects of Irradiation (1964)
1oc Radioactivity (1963)
10f Methods of Evaluating Radiological Equipment and Ma-
terials (1963)
12 Certification of Standardized Radioactive Sources (1968)
13 Neutron Fluence, Neutron Spectra and Kerma (1969)
15 Cameras for Image Intensifier Fluorography (1969)
16 Linear Energy nansfer (1970)
17 Radiation Dosimetry: X Rays Generated at Potentials of 5
to 150 kV(1970)
18 Specification of High Activity Gamma-Ray Sources (1970)
20 Radiation Protection Instrumentation and Its Application
(1970)
22 Measurement of Low-Level Radioactivity (1972)
Measurement ofAbsorbed Dose in a Phantom Irradiated
by a Single Beam o f X o r Gamma Rays (1973)
Determination of Absorbed Dose in a Patient Irradiated
by Beams o f X o r Gamma Rays in Radiotherapy Proce-
dures (1976)
Conceptual Basis for the Determination of Dose Equiva-
lent (1976)
Neutron Dosimetry for Biology and Medicine (1977)
A n International Neutron Dosimetry Intercomparison
(1978)
Basic Aspects of High Energy Particle Interactions and
Radiation Dosimetry (1978)
Quantitative Concepts and Dosimetry in Radiobiology
(1979)
Average Energy Required to Produce an Ion Pair (1979)
Methods ofAssessment ofAbsorbed Dose in Clinical Use
of Radionuclides (1979)
Radiation Quantities and Units (1980)
The Dosimetry of Pulsed Radiation (1982)
Microdosimetry (1983)
Stopping Powers for Electrons and Positrons (1984)
Dose and Volume Specification for Reporting Intracavi-
tary Therapy i n Gynecology (1985)
Determination of Dose Equivalents Resulting from Exter-
nal Radiation Sources (1985)
The Quality Factor i n Radiation Protection (1986)
Modulation 'IEansferFunction of Screen-Film Systems (1986)
Use of Computers i n External Beam Radiotherapy Proce-
dures with High-Energy Photons and Electrons (1987)
Determination of Dose Equivalents from External Radia-
tion Sources -Part 2 (1988)
%sue Substitutes i n Radiation Dosimetry and Measure-
ment (1989)
Clinical Neutron Dosimetry -Part I: Determination of
Absorbed Dose i n a Patient l k a t e d by External Beams
of Fast Neutrons (1989)
Photon, Electron, Proton and Neutron Interaction Data
for Body %sues (1992)
Photon, Electron, Proton and Neutron Interaction Data
for Body Tissues, with Data Disk (1992)
Measurement of Dose Equivalents from External Photon
and Electron Radiations (1992)
Phantoms and Computational Models in Therapy, Diag-
nosis and Protection (1992)
Stopping Powers and Ranges for Protons and Alpha
Particles (1993)
Stopping Powers and Ranges for Protons and Alpha
Particles, with Data Disk (1993)
Prescribing, Recording and Reporting Photon Beam
Therapy (1993)
Quantities and Units i n Radiation Protection Dosimetry
(1993)
Particle Counting i n Radioactivity Measurement (1994)
Gamma-Ray Spectrometry i n the Environment (1994)
Medical Imaging- The Assessment of Image Quality (1995)
Secondary Electron Spectra from Charged Particle Inter-
actions (1995)
 56 Dosimetry of External Beta Rays for Radiation Protection
(1997)
57 Conversion Coeficients for Use i n Radiological Protection
Against External Radiation (1997)
58 Dose and Volume Specification for Reporting Interstitial
Therapy (1997)
59 Clinical Proton Dosimetry -Part I: Beam Production,
Beam Delivery and Measurement ofAbsorbed Dose

Binders for ICRU Reports are available. Each binder will accommodate from six to eight reports. The binders
carry the identification, "ICRU Reports", and come with label holders which permit the user to attach labels
showing the Reports contained in each binder.
The following bound sets of ICRU Reports are also available:
Volume I. ICRU Reports lob, 10c, 10f
Volume 11. ICRU Reports 12,13,14,15,16,17,18,20
Volume 111. ICRU Reports 22,23,24,25,26
Volume N. ICRU Reports 27,28,30,31,32
Volume V. ICRU Reports 33,34,35,36
Volume VI. ICRU Reports 37,38,39,40,41
Volume VII. ICRU Reports 42,43,44
Volume VIII. ICRU Reports 45,46,47
Volume M. ICRU Reports 48,49,50,51
(Titles of the individual Reports contained in each volume are given in the list of Reports set out above.)
The following ICRU Reports were superseded by subsequent Reports and are now out of print:

ICR U
Report No. Title and Reference*
1 Discussion on International Units and Standards for
X-ray work, Br. J . Radiol. 23,64 (1927).
2 International X-Ray Unit of Intensity, Br. J . Radiol. (new
series) 1,363 (1928).
3 Report of Committee on Standardization of X-ray Mea-
surements, Radiology 22,289 (1934).
4 Recommendations of the International Committee for Ra-
diological Units, Radiology 23,580 (1934).
5 Recommendations of the International Committee for Ra-
diological Units, Radiology 29,634 (1937).
6 Recommendations of the International Commission on
Radiological Protection and of the International Com-
mission on Radiological Units, National Bureau of
Standards Handbook 47 (U.S. Government Printing
Office, Washington, D.C., 1951).
7 Recommendations of the International Commission on
Radiological Units, Radiology 62,106 (1954).
8 Report of the International Commission on Radiological
Units and Measurements (ICRU) 1956, National Bu-
reau of Standards Handbook 62 (U.S. Government
Printing Office, Washington, D.C., 1957).
9 Report of the International Commission on Radiological
Units and Measurements (ICRU) 1959, National Bu-
reau of Standards Handbook 78 (U.S. Government
Printing Office, Washington, D.C., 1961).
Radiation Quantities and Units, National Bureau of
Standards Handbook 84 (U.S. Government Printing
Office, Washington, D.C., 1962).
Clinical Dosimetry, National Bureau of Standards Hand-
book 87 (U.S. Government Printing Office, Washington,
D.C., 1968).
Radiobiological Dosimetry, National Bureau of Standards
Handbook 88 (U.S. Government Printing Office, Wash-
ington, D.C., 1963).
Radiation Quantities and Units (International Commis-
sion on Radiation Units and Measurements, Washing-
ton, D.C., 1968).
Radiation Dosimetry: XRays and Gamma Rays with
Maximum Photon Energies Between 0.6 and 50 MeV
(1969).
Radiation Quantities and Units (International Commis-
sion on Radiation Units and Measurements, Washing-
ton, D.C., 1971).
Dose Equivalent [Supplement to ICRU Report 191(Inter-
national Commission on Radiation Units and Measure-
ments, Washington, D.C., 1973).
Radiation Dosimetry: Electrons with Initial Energies Be-
tween 1 and 50 MeV (International Commission on Ra-
diation Units and Measurements, Washington, D.C.,
1972).
Dose Specification for Reporting External Beam Therapy
with Photons and Electrons (International Commission
on Radiation Units and Measurements, Washington,
D.C., 1978)
Radiation Dosimetry: Electron Beams with Energies Be-
tween 1 and 50 MeV (International Commission on Ra-
diation Units and Measurements, Bethesda, MD,
1984).

*Reference given are in English. Some of the Reports were also pub-
lished in other languages.
 Index
Absorbed dose, 11,12,13,15,17,18,19,26,27,37,41,44 Commission of a treatment system, 33
Based on absolute dose to water calibrations, 27 Measurements, 33
Based on air kerma calibrations, 27 Common dosimetric materials, 20
Based on fluence, 26 Composition of dosimetric materials, 20
Absorbed dose as a function of depth, 15 Considerations of w and W, 21
Absorbed dose determination, 18 Contoured second scatterer, 9
Absorbed dose in gas, 19 Cyclotrons, 6
Absorbed dose in a phantom, 37
Absorbed dose measurements, 33 Daily checks, 43
Absorbed dose measurements with a calorimeter, 17 Delta ray, 13
Absorbed dose measurements with an ionization chamber, 18 Depth-dose curves, 2
Absorbed dose rate, 11 Depth dose distribution, 1,35
Absorbed dose to water, 26,27,39 Depth dose measurements, 35
Absorbed dose, 26 Depth of measurement, 37
Based on absolute dose to water calibrations, 27 Detectors, 29
Based on air kerma calibrations, 27 Detectors for dose distribution measurements, 33
Based on fluence, 26 Detedors for relative dosimetry, 33,34
Determination of absorbed dose to water, 26 Determination of absorbed dose in a phantom, 37
Absorbed dose to water using a calibrated ionization chamber, 37 Determination of absorbed dose to water, 26,27
Absorbed dose to water using a calorimeter, 39 Based on absolute dose to water calibrations, 27
Absorbed dose to water using a faraday cup, 40 Based on air kerma, 27
Accelerators, 6 Based on fluence, 26
Alanine detectors, 34 Determination of dose distributions, 35
Angle decoder, 29 Determination of proton absorbed dose, 41
Average energy loss, 13 Determination of proton absorbed dose in reference conditions,
Average energy required to produce an ion pair, 13 15
Determining the absorbed dose in water using an ionization
Beam characteristics, 14 chamber, 44
Beam centering devices, 29 Diamond detedow, 35
Beam delivery techniques, 8 Discrete scanning,10
Dynamic, 8 Dose deposition of the protons, 35
Passive, 8 Dose distributions, 33,35
Beam energy monitors, 29 Dose measurements, 33
Beam intensity monitors, 29 Dose monitoring, 29
Beam monitoring, 29 Dose uniformity, 36
Beam monitoring detectors, 29,32 Dosimeter calibration, 17
Beam monitoring precision, 32 Dosimetric quantities, 11,12
Beam penumbra, 36 Dosimetry measurement systems, 33
Beam position monitors, 29 Dosimetry of proton beams, 37
Beam position monitoring, 32 Dosimetry worksheet, 44
Beam profile monitors, 29 Double scattering, 8
Beam range, 35 Dynamic beam delivery, 8 , 9 , 1 0
Beam scanning, 9 Advantages, 10
Beam shaping and delivery, 8 Disadvantages, 10
Beam spreading, 9 Discrete Scanning, 10
Biannual checks, 43 Organ motion, 10
Bolus, 9 , 1 0 Patient motion, 10
Bragg peak, 1 Raster scanning, 9
Spot scanning, 10
Calibrated ionization chamber, 37 Voxel scanning, 1
Calibration factor, 37 Wobbling, 9
Calorimeter, 17 Dynamic beam delivery monitoring, 32
Calorimetric absorbed dose determination, 17 Drift distance, 9
Calorimetric measurement, 17
Calorimetry, 18 Effective stopping power, 19
Technical problems, 18 Electromagnetic interactions, 13
Cema, 12,13,15,18,19,26 Electron secondary-emission monitor, 32
Cema approximation, 16 Elemental composition of several solids, liquids and gases, 20
Charge per proton produced by traversal, 30 Elemental compositions of several important dosimetric materi-
Chemical dosimeters, 35 als, 20
Choice of dosimetric instruments, 37 Energy absorption event, 12
Choice of values to be used for physical parameters, 37 Energy deposition, 13
Clinical dosimetry measurements, 33 Energy fluence, 11
Collection fraction, 31 Energy fluence rate, 11
Energy flux, 11 New treatment portals, 43
Energy modulation, 1 Non-stochastic quantities, 11
Energy straggling, 13 Nuclear interactions, 13,14
Numerical values of required quantities, 40
Facilities, 2, 4 Numerical values of physical parameters, 41
Factors which affect proton beam characteristics, 14 &based calibrated ionization chamber, 37
Faraday cup, 16,17,40 Nw-based calibrated ionization chamber. 39
Faraday cup calibration, 26
Field size, 36 Operating Parameters, 8
Film response, 34 Beam current, 8
Films Energy, 8
Radiographic film, 34 Range, 8
Fixed range modulation, 9 Time structure, 8
Fluence, 40 Treatment time, 8
Fluence measurements with a faraday cup, 16
FM cyclotron, 6 Pad ionization chambers, 32
Fraction of ions collected in air versus proton fluence rate, 32 Particle flux,11
Frequency modulated cyclotron, 6 Particle fluence, 11
Frequency of check procedures, 43 Particle fluence rate, 11
Passive beam delivery, 8
Gas ionization techniques, 18 Passive scattering techniques, 8 , 9
Gray, 11 Contoured second scatterer, 9
Disadvantages, 9
History of proton therapy, 2 Double scattering, 8
Magnetic dispersion, 9
Intensity monitors, 29 Patient position monitors, 29
Interaction coefficients, 11 Penumbra widths, 36
Interpretation of ionization chamber response, 19,21 Penumbra characteristics, 36
Consideration of w and W, 21 Periodic quality assurance procedures, 43
Effective stopping power approximation, 19 Phantom composition, 33
Stopping power, 19 Phantoms, 35
Introduction, 1 Phantom material, 37
Ion collection fraction, 31 Physical parameters, 41
Ion recombination effects, 31 Physical quantities, 11
Ionization chamber, 18,37 Pion treatments, 3
Ionization chamber response, 19,21 Plateau, 1
Consideration of w and W, 21 Positmn camera, 29
Effective stopping power approximation, 19 Precision of monitoring of dynamic beam delivery, 32
Stopping power, 19 Production of proton beams for therapeutic applications, 6
Ionization chambers for beam monitoring, 29 Proton accelerators, 6
Irradiation medium for clinical dosimetry, 33 Proton beam characteristics, 14
Ionization methods of measurement of absorbed dose, 18 Proton calibration factor, 37
Ionization produced per proton versus proton energy, 30 Proton csda ranges, 23
Isochronous cyclotron, 6 , 7 Proton dosimetry, 15, 16
General considerations, 15
Jesse effect, 30,31 Geometry of the detector, 16
Material of interest, 16
Lateral dose uniformity, 36 Proton interactions with matter, 13
LET, 13 Proton mass electronic stopping powers, 38
Lineal energy, 12 Proton mass stopping power, 13
Lineal energy distribution, 14 Proton therapy facilities, 4
Linear accelerators, 8
Linear energy transfer, 12,13 Quality assurance, 43
Quantities, I1
Machine reliability, 8
Machine stability, 8 Radiochromic film, 35
Magnetic dispersion, 9 Radiometric quantities, 11
Mass electronic stopping power, 11,14,38 Range, 23
Mass electronic stopping power values, 22 Range measurements, 35
Mean energy expended in a gas per ion pair formed, 12 Range of particle, 13
Measurement of beam range, 35 Range straggling, 13
Microdosimetric concepts, 14 Raster scanning, 9
Modulation, 1 , 2 Rate quantities, 11
Monitoring, 29 Ratio of the charge per proton produced by traversal, 30
Monitoring beam intensity, 29 Rationale for Proton Therapy, 1
Monitoring beam profile, 29 Ratios of the mass electronic stopping power of several materials
Monitoring precision, 32 to air, 20
Monitoring of dynamic beam delivery, 32 Ratios of the mass electronic stopping power of several materials
to TE-methane, 21
Negative pions, 3 RBE, 2
Recombination, 30 Superconducting cyclotrons, 7
Recombination effects, 31 Superconducting synchrocyclotron, 7
Recommend numerical values of parameters, 40 Synchrotrons, 7
Recommendations, 41 Synchrocyclotrons, 6
Recommendations for determination of absorbed dose in a phan-
tom, 37 Tests, 43
Reference depth, 37 Thimble ionization chambers with a =OCocalibration factor, 37
Reference dosimeter, 37 Time structure of the beam, 33
Reference point for calibration, 37 TLDs, 35
References, 46 Total absorbed dose a t depth, 15
Relation to existing reports, 5 Transmission ionization chamber, 30
Relative abcorbed dose, 15 Treatment facilities, 2
Relative absorbed dose as a function of depth, 15 Treatment times. 8
Relative biological effectiveness, 2
Relative dose measurements, 33 Uncertainty in calorimetric dose determination, 40
Relative dosimetry, 29,30,33 Uncertainty in the absorbed dose determined by ionization
Relative proton and neutron spectral fluence per incident proton, chamber dosimetry, 39
16 Units, 11
Residual csda range, 41 Uniformity characteristics, 36
Response of film, 34
Uniformity of dose, 36
Response of silicon diodes vs. ionization chambers, 34
Restricted linear collision stopping power, 12
Values of the physical parameters, 40,41
Restricted stopping power, 13
Verifymg the distribution of dose within the treatment volume,
Ridge filter, 9
33
Voxel scanning, 10
Scintillator cell, 32
Scope of report, 4
Secondary-emission monitors, 32 w, 21
Semiconductor diodes, 33 W, 12,13,21
Silicon diodes, 33 w values, 23,25
Silicon diode response, 34 Wle, 24
Si-diode and parallel plate ionization chamber response, 34 Wle values, 21,24,25,26
SI, 11 wle values, 23,25,26
SOBP, 1 Wall effects, 39
Specificenergy, 12,14 Water calorimeters, 18
Spot scanning, 10 Water-equivalent path lengths, 35
Spread-out Bragg peak, 1 Weekly checks, 43
Standard reference medium, 33 Weighting factor, 2
Stochastic quantities, 11,12 Wobbling and beam scanning, 9
Stopping power, 19,22,38 Worksheet, 44
Strip ionization chambers, 32
Summary of recommendations, 41 Yearly checks, 43