Radiation Protection Dosimetry (2012), Vol. 149, No. 3, pp.

297 303
Advance Access publication 7 July 2011

doi:10.1093/rpd/ncr254

EFFECTIVE DOSE FOR SCOLIOSIS PATIENTS UNDERGOING

FULL SPINE RADIOGRAPHY
M. Mogaadi1,2,*, L. Ben Omrane2 and A. Hammou2
1
Department of Biology, University of Bizerte, Bizerte
2
National Center of Radiation Protection, Tunis, Tunisia

Received September 16 2010, revised May 28 2011, accepted June 2 2011

Scoliotic patients underwent many radiological examinations during their control and treatment periods. Nowadays, few
studies have calculated effective dose which is the primary indicator of radiation risk. In this study, the PCXMC program is
used to calculate the effective doses associated with scoliosis radiography. Five age groups of patients, proposed by the
National Radiological Protection Board, have been chosen: <1, 1 4, 5 9, 10 15 and 16 y (adult patients). Patient and
radiographic data were collected from 99 patient examinations for both anteroposterior and lateral full spine X-ray projections. Results showed the effective dose ranged from 118 to 1596 mSv for the frontal projection and from 97 to 1370 mSv for
the lateral projection, with patient age varying from 3 months to 22 y. This study presents the effective dose against patient
age and demonstrates the necessity to optimise patient protection for this type of examination.

INTRODUCTION
Scoliosis is defined as a lateral curvature of the spine;
the deformity is more complicated as it is not limited
to one plane only. A lordosis rotates to one side, producing scoliosis as a secondary phenomenon; there
is, therefore, deformity in all three planes(1, 2).
Initial diagnosis of scoliosis is usually made by
physical examination using the forward bending or
Adamss test. Although, for a definitive diagnosis,
some doctors then use a scoliometer system to
measure the curvature if the child is thought to have
scoliosis, a radiographic examination is usually done,
which should include standing, frontal and lateral
views of the spine. Radiological imaging is necessary
to clarify the nature of scoliosis and the impact of
spine deformation on other systems(3). The frontal
projection is performed both in the anteroposterior
(AP) or posteroanterior (PA) position. This projection is crucial in determining the degree of distortion
and useful to measure the angle of curvature of the
spine according to Cobbs method. The lateral projection is very useful especially during the first investigation to reveal the kyphosis type of deformation(2).
The most observed scoliosis patients are children.
They constitute a population group with a high sensitivity to ionising radiation(4), and they undergo
many examinations during their control and treatment periods. Therefore, this group of patients was
chosen in order to carry out a detailed survey and to
compare the results of this study with those published from other countries.
As the received doses during this examination are
not at the level to induce deterministic effects, the
effective dose (E) has been assumed. E is an

important variable because it takes into account the

sensitivity of different organs and tissues(4) and is
the primary indicator of radiation risk of
malignancy.
In this study, the effective dose has been calculated
with a computer program called PCXMC. The
reliability of this software has been proved by the
work released by Servomaa and Tapiovaara(5). Using
Monte Carlo simulation, the X-ray beam field size
and position must be related to the mathematical
phantom. The beam penetration parameters, such as
peak kilovoltage and total filtration must be known
as well as the entrance air kerma (EAK).Various
ages, heights and weights, in widely adjustable projections and other conditions of X-ray imaging, are considered by this program.
Comparison of the results here with those from
other works was also presented. The exposure parameters and patients weight and age were fully
investigated.

MATERIALS AND METHODS

All the examinations considered in this study were
undertaken in a childrens hospital. The exposures
were performed using a high-voltage (HV) generator
Villa type Genius 7503 from Villa Sistemi Medicali
manufacturer. The X-ray tube is IAE, Type 101H
RTM (manufacturer IAE spa), with a nominal large
focal spot of 1.5 mm and the target angle of 168.
The total filtration was estimated to be 2.5 mm of
aluminium (Al) based on the half-value layer
(HVLs) measurement (6).

# The Author 2011. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Downloaded from http://rpd.oxfordjournals.org/ at Stefan cel Mare University from Suceava on December 12, 2014

*Corresponding author: mogaadihama@yahoo.fr

M. MOGAADI ET AL.

mathematical phantom used by the program. The

final stage of the PCXMC analysis involves the calculation of organ and effective doses using the spectrum characteristics ( peak kilovoltage, total filtration
and angle of anode) and the EAK.
To analyse the results of the effective doses from
this radiology examination, a distribution of age
proposed by the National Radiological Protection
Board (NRPB)(8) was applied to the patient sample:
newborns, infants (0 1 y), 1 4, 5 9, 10 15 y and
adults (16 y).
RESULTS
X-ray technique
Figure 1 shows the radiation output (mGy mA s21)
against X-ray tube potential at a distance of 100 cm
from the focal spot.
Table 1 provides a statistical summary of the different characteristics of the 99 patients who underwent
a scoliosis examination. Most of the patients were
children (86 %) with good dispersal through all age
intervals. Their age ranged between 3 months and 22
y. According to this case, the patients belong to
different periods of childhood and the weight varied
by a factor of 11 (775 kg), and the thickness varied
by about a factor of 3 in the AP view and of 2 in the
lateral view. This fluctuation in the body size and age
led to a considerable variation in the parameters used
in the AP and lateral projections.
For the frontal view, for children the X-ray tube
potentials ranged between 57 and 86 kVp with milliamperes per second values varying from 16 to 150

Figure 1. Measured X-ray tube output (mGy mA s21) at a

distance of 100 cm against X-ray tube potential (kVp).

298

Downloaded from http://rpd.oxfordjournals.org/ at Stefan cel Mare University from Suceava on December 12, 2014

The radiological examinations of the full spine

with frontal and lateral views were most often performed in a standing position with a focus film distance (FFD) of 185 cm. In some cases a lying
position was used (infants and recently operated
patients) with an FFD of 115 cm. All frontal projections are performed in an AP position. Two kinds of
cassettes (smaller cassette: 3643 cm2 and larger
cassette: 30120 cm2) were used. The smaller cassette was routinely adopted for smaller scoliotic children, generally put on to the Bucky tray. The
screenfilm combination was AGFA ortho400/
AGFA orthomatic G-plus (400-speed screen-film
system). Most scoliosis X-ray examinations, however,
use a large cassette which was specially designed for
this type of radiographic examination, and gave an
image of the combined cervical, thoracic and
abdominal spine regions. Three films (AGFA orthomatic G-plus) of 3040 cm2 were put in this cassette
containing an intensifying screen (CAWO G
gradual2/) where the speed gradually decreases
across from edge to edge, and was indicated on the
screen by and 2 signs(7) (speed class: 100
400). A specially adapted range of speed compensates high differences in density of body.
Radiographic exposure factors were selected by a
radiographer. Patient sex, age, weight, height, thickness, tube potential, milliamperes per second, FFD
and exposed film area were recorded for each
radiograph.
A quality control check was performed for the full
installation before the patient measurements. The Xray beam HVLs was 3.26 mm of Al measured at 80
kVp using a Victoreen 8000 detector.
The X-ray tube output was measured at 100 cm
from the focus for HV ranged between 60 and 90
kVp with an ionising chamber NE-2571 calibrated at
the International Atomic Energy Agency Dosimetry
Laboratory. The overall uncertainty associated with
ionisation chamber measurements was determined to
be 5.5 % (for 95 % confidence level). These data were
used to extrapolate EAK for corresponding exposure
factors ( peak kilovoltage and milliamperes per
second) at the focus skin distance for each radiographic projection in the study. The EAK without
backscatter factor was then used as an input for effective dose calculation for each projection using the
Monte Carlo program PCXMC(5).
The
PCXMC
Monte
Carlo
program
(PCXMC1.5.1, STUK, Radiation and Nuclear
Safety Authority, Helsinki, Finland) was used to calculate the effective dose from the experimentally
measured EAK. PCXMC uses hermaphrodite mathematical phantoms based on the models of Cristy
(1980), which are adjustable for weight, height and
ages of 1, 5, 10, 15 y and adult. A critical aspect of
the method was the need to accurately describe
the X-ray beam field size and position related to the

EFFECTIVE DOSE UNDERGOING FULL SPINE RADIOGRAPHY

Table 1. Statistical summary of patient characteristics.
Range of age (y)

Age (y)a

Weight (kg)

Height (cm)

AP thickness (cm)

Lateral thickness (cm)

,1 (n6)
1 4 (n21)
5 9 (n19)
10 15 (n40)
16 (n13)
Total number99

0.6 (0.250.8).
2.8 (1.1 4.6)
7.1 (59.7)
13.2 (10 15.7)
17.6 (16 22)
9.7 (0.2522)

8 (79)
12 (817)
22 (13 35)
38 (20 75)
48 (22 65)
29 (775)

69 (6274)
89 (71108)
120 (92141)
148 (110 173)
158 (125 182)
127 (62182)

11.6 (10 15)

12.7 (10 15)
15.6 (12 20)
17.9 (12 27)
20.9 (15 27)
16.4 (10 27)

14.4 (1415)
15.9 (1418)
19.2 (1423)
22.6 (1729)
24 (1827)
20.2 (1429)

Table 2. Mean values of weight, EAK and effective dose E

in each age group for frontal and lateral projections.
Age groups (y)

Weight (kg)
EAK (mGy): frontal
EAK (mGy): lateral
E (mSv): frontal
E (mSv): lateral

0 1

1 4

5 9

10 15

16

8
479
796
229
169

12
589
967
299
227

22
830
1495
450
358

38
1275
2733
678
586

48
1495
3112
798
597

mA s, and for adult patients, the examination parameters ranged between 71 and 81 for peak kilovoltages and between 80 and 150 for mA s.
For the lateral view, for children the X-ray tube
potentials ranged between 58 and 97 kVpwith milliamperes per second values varying from 16 to 240
mA s, and for adult patients, the peak kilovoltage
ranged between 73 and 98 and the mA s between
120 and 240.
The beam area at the patient surface varied significantly (10 times) for the whole patient sample
(from 291 cm2 to 3114 cm2). The mean dimensions
of the entrance X-ray beam was (2358 cm2) either
for the lateral or the frontal views. A poor collimation was undertaken by the radiographers, which
tend to cover the full spine, exceeding sometimes the
cassette size, especially if the direction of deformation was unknown.
Patient doses
An overview of the EAK measurements for full
spine frontal and lateral examinations is presented in
Table 2. Values for lateral projection are higher than
those for frontal projection in all age groups.
The average values of effective doses for frontal
and lateral views are also presented in Table 2 for
each age group. According to the results in this
table, the effective dose to patients in 10 15 y of age
was three and four times higher than the infant

(1 12 months) dose for frontal and lateral views,

respectively. This result confirmed the influence of
patient size on effective dose received in diagnostic
radiology(9, 10). The E values for adults, with an
average weight of 48 kg, for frontal and lateral views
were little higher than for children aged from 10 to
15 y with an average weight of 38 kg because the
body size of scoliotic patients in these two groups
was similar.
The values of effective dose are higher for the
frontal view than those for the lateral view, while the
exposure parameters (peak kilovoltage and milliamperes per second) are higher in lateral incidence. The
X-ray beam intercepts directly most of the sensitive
organs in the frontal view than in the lateral view.
Figure 2 shows the values of patient effective dose
as a function of patient age for frontal and lateral
views. An acceptable correlation was found between
these two parameters, both for the frontal or lateral
views. The general shape of the curve shows that the
effective dose tends to increase with the patient age.
However, by considering the scatter plot, three
regions appeared: the first one for children patients
aged ,3 y with the points are closer to the curve
and the second one for patients who were aged
between 3 and 7 y with the points slightly dispersed
and the last one for children .7 y old with a large
spread of points around the fitted curve.
Figure 3 shows the computed values of patient
effective dose according to patient weight in frontal
and lateral projections. Table 3 provides a summary
of the fitted data relating the effective dose E to the
patient weight, together with the coefficient of determination (r2), which shows an acceptable correlation
between these two parameters if the whole sample of
patients is considered. Between 30 and 75 kg, the
correlation becomes weak and then increases slightly
for patients who weigh between 20 and 75 kg. These
factors were similar to those reported in other
studies(11, 12) by considering the same range of
patients weight. On the other hand, by reducing the
interval of patient weight, from the total patient
sample to intervals of patients having a weight of
.20 kg then 30 kg, the patient characteristics

299

Downloaded from http://rpd.oxfordjournals.org/ at Stefan cel Mare University from Suceava on December 12, 2014

n, number of patients.
a
First figure is the mean and the range is given in parentheses.

M. MOGAADI ET AL.

Figure 3. Effective dose E against patient weight for frontal and lateral views.

300

Downloaded from http://rpd.oxfordjournals.org/ at Stefan cel Mare University from Suceava on December 12, 2014

Figure 2. Effective dose E against patient age for frontal and lateral views.

EFFECTIVE DOSE UNDERGOING FULL SPINE RADIOGRAPHY

Table 3. Coefficient of determination (r2) between effective dose (E) and patient weight (W) for the frontal and lateral
projections and weight variation, for the different selection of patients weight and for the BMI patients selection.
Full spine examination
Number of patients
Weight variation
(maximum/minimum)
(r2): frontal
(r2): lateral

7 kg , W , 75kg
99
11
0.67
0.64

W . 20 kg

W . 30 kg

36
3.75

42
2.5

0.27
0.27

DISCUSSION
The X-ray output (mGy mA s21) depicted in
Figure 1 is three times higher than that presented by
Chamberlain et al. (11) because a wedge-shaped aluminium filter is not used in this study and the total
filtration is thin (2.5 mm Al). The filtration of the
primary X-ray beam reduces the total number of
X-ray photons in the beam but, more importantly, it
selectively removes a greater portion of the lowenergy photons that do not contribute to the
production of a radiograph. For the same reasons,
EAK values summarised in Table 2 are higher than
those found in the literature(11, 13).
Generally, effective dose tends to increase with the
patient age, but the dispersion depicted in Figure 2
which is due to the large variation of patient size
confirms the age distribution recommended by
NRPB. By considering the scatter regions observed
in this figure, the International Commission of
Radiation Protection age groups (3 months, 1, 27,
7 12, 12 17 y) seem to be more adapted.
Contrary to what is expected with the results of the
good correlation between effective dose and patient
weight, a limited dependence on the patient size is
observed especially for a weight of .30 kg which corresponds to child age of nearly .10 y having a slow
growth compared with children having an age of ,10
y. On average, effective dose tends to increase with
patient weight, but the contribution at the same time
of other factors such as the patients age and size do
not permit any definitive conclusions to be reached.
The question however still remains with regard to
the reservation on the use of the PCXMC program
for this special case of patients. To investigate the
body mass index (BMI weight/height2), which
define the body fatness in medicine was derived for
the sample in whom 35 % of them fall outside of the
normal expectations of size. While the remaining
patients are generally underweight, it is seen in
Table 3 that the coefficient of determination between
effective dose and patient weight remains unchanged

65
10
0.68
0.67

or improves slightly when the BMI selection is

applied to the sample. Despite these results, caution
should be taken that this program should not be
used for individual cases for estimation of effective
dose, but over a sample of population should yield
valuable dosimetric information and encourage to
dose optimisation.
The comparison of effective dose is difficult in
paediatric radiography because of the wide variation
in children size and radiographic techniques used in
full spine examination.
Table 4 outlines the results from four previous
studies(11 14) compared with the present study. The
values found in this work were higher than those
published by other authors either for paediatric
studies or for adult ones. The most important
factors responsible for this increase could be divided
into two categories: the first relevant to the technology of radiological equipment and the second due
essentially to the radiological practice.
The speed of the screen film system is one of the
most critical factors affecting the radiation dose to
the patient (15). As noticed in this table, the speed
index of this study was lower compared with similar
studies. The European Commission (EC) guideline
recommends a nominal speed class of 600800(16).
On the other hand, the filtration is a factor which
is no less important for the high doses encountered
in this study. Most X-ray tubes have a minimum
total filtration of 2.5 mm Al. The EC guideline(16)
stated that an additional filtration up to 1 mm Al
plus 0.1 mm or 0.2 mm copper (Cu) could be appropriate for paediatric patient.
Very often, wide fields were used by the radiographers to cover the full spine curvature, inducing
unnecessary exposure of some organs. It was also
observed that, high milliamperes per second values
were used, for example, as seen in Table 4, on
frontal projection a mean value of 70 mA s was
noticed for children while for similar works(12) the
mean milliamperes per second value was nearly
50 % lower. By considering the same reference, the
mean milliamperes per second values in this study
were three times higher for the lateral incidence.
In an optimisation context, the radiographic protocol has been investigated. It is seen in Table 4 that

301

Downloaded from http://rpd.oxfordjournals.org/ at Stefan cel Mare University from Suceava on December 12, 2014

becomes more homogenous in terms of weight

variability (from 11 to 2.5 times) and consequently
the correlation factor decreases.

0.16
0.18

BMI selection

M. MOGAADI ET AL.
Table 4. Effective dose E in scoliosis radiography for ulterior studies.
Study

Number of patients (age)

E: mean
(minimum maximum)

Speed class

600
800
600

Gerjer et al.(14)
This study

400
Gradual speed

PA and lateral: 120 mSv

AP: (140180) mSv
PA: (81 123 mSv)
Lateral: (124207 mSv)
Female Alderson phantom PA: 160 mSv
86 (3 months 15 y)
AP: 503 mSv
(118 1596 mSv)
Lateral: 421 mSv
(971370 mSv)
PA: 252 mSv (56822 mSv)
13 (16 22 y)
AP: 798 mSv
(545 1305 mSv)
Lateral: 597 mSv
(344 1187 mSv)
PA: 422 mSv (294 674 mSv)

(100400)

54 (27 95)
33.9 (18 50)
39.9 (25 50)
40
69.6 (16 180)
116.9 (16 320)
108.7 (80 150)
184.7 (120240)

Min, minimum; max, maximum.

the effective doses calculated for the same sample

and the same conditions could be reduced by a
factor of 2 simply by turning the patient and
taking a posterioanterior (PA) view, which is
recommended by WHO(13, 17). Effectively, the AP
projection reduces image magnification and helps to
minimize focal spot and motion blur but, at the
same time increases the effective dose as most sensitive organs are more towards the anterior side.
Indeed the results of this work show that it is
necessary to refine the technical parameters and the
radiological protocol to perform this type of examination in this hospital and to use adequate materials
such as the screen-film system and the filtration.
It is of interest to notice that during childhood,
scoliotic patients generally undergo one full spine
examination each year as a control of the curvature.
But during the age of puberty they will have
approximately 22 radiological examinations over a
3-y treatment period. This results obviously in a
cumulative dose of 9 mSv y21 which is nearly eight
times the natural background level, normally taken
to be 1.12 mSv y21. Considering further special
investigations, such as tomodensitometry, these
patients may receive at least twice the further cumulative dose each year during the puberty period.
Special attention should be directed then to this type
of examination.
CONCLUSION
The effective dose values calculated with the
PCXMC program in full spine examinations in this
study are four times higher than results of other
authors. Nevertheless, an important potential for
dose reduction has been noted, in the radiological

protocol with the PA incidence, in the examination

parameters selected by the radiographers with an
adaptation to patient size and in the technology
affecting the image quality like the screen films and
the tube filtration. It is necessary to optimize the
system for performing radiographic examinations of
children in the hospital and encourage radiographers
to use good collimation.
Variations in size are marked not only for adults
but also for paediatric patients and the use of a
single reference size is impractical. Accordingly, the
use of many age groups is useful to estimate the
effective dose and to control the patient exposure
during the whole treatment period.

ACKNOWLEDGEMENTS
The authors would like to acknowledge the statistical advice from Dr. Mossadok Ben Attia at the
University of Bizerte and the assistance of the staff
of National Center of Radiation Protection of Tunis
and the staff of radiological service of Childrens
Hospital of Tunis.

FUNDING
This work has received funding from the Research
Unit of the National Center of Radiation Protection
of Tunis.
REFERENCES

302

1. Deacon, P., Archer, I. A. and Dickson, R. A. The

anatomy of spinal deformity: a biomechanical analysis.
Orthopedics 10, 897 903 (1987).

Downloaded from http://rpd.oxfordjournals.org/ at Stefan cel Mare University from Suceava on December 12, 2014

7 (925 y)
Hallen et al.(13)
Chamberlain et al.(11) 61 (9 55 y)
54 (4 17 y)
Chih-l et al.(12)

mA s: mean
(minimum maximum)

EFFECTIVE DOSE UNDERGOING FULL SPINE RADIOGRAPHY

11.

12.
13.
14.

15.

16.

17.

303

phantoms of varying voxel size. Radiat. Prot. Dosim.

49, 353356 (1993).
Chamberlain, C. C., Huda, W., Hojnowski, L. S.,
Perkins, A. and Scaramuzzino, A. Radiation doses to
patients undergoing scoliosis radiography. Br. J. Radiol.
73, 847853 (2000).
Chih-l, L., Donald, M. C. and John, R. Measurement
of effective dose for paediatric scoliotic patients.
Radiography 11, 89 97 (2005).
Hallen, S., Martling, K. and Mattsson, S. Dosimetry at
X-ray examinations of scoliosis. Radiat. Prot. Dosim.
43, 49 54 (1992).
Geijer, H., Beckman, K.-W., Jonsson, B., Andersson,
T. and Persliden, J. Digital radiography of scoliosis with
a scanning method: initial evaluation. Radiology 218,
402410 (2001).
Hart, D. and Wall, B. F. A study of causes of variations
in doses to patients from X-ray examinations: a report
to the Department of Health, NRPB M212. National
Radiological Protection Board (1990).
European Commission. European guidelines on quality
criteria for diagnostic radiographic images in paediatric.
Pub-lication EUR16261 EN. European Commission
(1996).
WHO. Rational use of diagnostic imaging in paediatrics
report of a study group. World Health Organisation
Technical Report series 757. WHO (1987).

Downloaded from http://rpd.oxfordjournals.org/ at Stefan cel Mare University from Suceava on December 12, 2014

2. Deacon, P., Flood, B. M. and Dickson, R. A.

Idiopathic scoliosis in three dimensions. A radiographic
and morphometic analysis. J. Bone Joint Surg. Br. 66,
509 512 (1984).
3. Bernard, J. C., Schneider, M., Biot, B. and Jemni, S.
Imagerie du scoliotique. Resonances Europeennes du
Rachis. Volume 13 N8 41, 1671 1675 (2005).
4. International Commission on Radiological Protection.
1990 recommendations of International Commission on
Radiological Protection. Publication 60. Ann. ICRP 21,
1 201 (1991).
5. Servomaa, A. and Tapiovaara, M. Organ dose calculation in medical X ray examinations by the program
PCXMC. Radiat. Prot. Dosim. 80, 213219 (1998).
6. Cranley, K. and Fogarty, G. W. A. The measurement of
total filtration of diagnostic X-ray tubes. Br. J. Radiol.
61, 388392 (1988).
7. John, B. and Tony, P. Chesneys Radiographic Imaging,
sixth edn, Blackwell Science, pp. 5565 (1995).
8. Hart, D., Wall, B. F., Shrimpton, P. C., Bungay, D. R.
and Dance, D. R. Reference doses and patient size in
paediatric radiology, Chilton,NRPB-R318 (2000).
9. Veit, R. and Zankl, M. Influence of patient size on
organ doses in diagnostic radiology. Radiat. Prot.
Dosim. 43, 241243 (1992).
10. Veit, R. and Zankl, M. Variation of organ doses in paediatric radiology due to patient diameter, calculated with