BASUMA – The Sixth Sense for Chronically Ill Patients
Thomas Falck, Javier Espina Jean-Pierre Ebert, Daniel Dietterle
Philips Research Laboratories Aachen
{Thomas.Falck, Javier.Espina}@philips.com
Abstract
Continuous monitoring and analyzing of vital signs
is the key for detecting at an early stage when a pa-
tient’s state of health changes to the worse, thereby
preventing emergency cases, which are harmful to the
patient and costly for the healthcare system. The
BASUMA project is concerned with developing an en-
ergy-efficient and robust system-on-chip platform for
wireless body sensors networks that enable health
monitoring of chronically ill patients in their own
homes. Initial application areas of BASUMA are: im-
proving the treatment of chronic obstructive pulmo-
nary disease patients and enhancing the ambulatory
chemo therapy of women suffering from breast cancer.
1. Introduction
The exploding costs of managing patients with
chronic conditions (e.g. cardio-vascular diseases,
chronic obstructive pulmonary disease, asthma, diabe-
tes and dementia) require new ways to support these
patients more effectively in a homecare setting.
Several ongoing European projects are addressing
the need for enhancing the quality-of-care while reduc-
ing health expenditures by developing mobile health
monitoring systems.
For example, the HealthService24 [1] project aims
to offer a viable mobile healthcare service permitting
health professionals to remotely assess, diagnose and
treat patients whilst the patients stay fully mobile. Their
platform is centered on a mobile phone that acts as
gateway for the wireless body-worn sensors. The mo-
bile phone constantly forwards the collected data via
UMTS to a medical service center or to medical pro-
fessionals for processing the vital signs and providing
personalized feedback to the patient.
In contrast the MyHeart [2] project is committed to
fight cardio-vascular diseases by developing intelligent
Proceedings of the International Workshop on Wearable and Implantable Body Sensor Networks (BSN’06)
0-7695-2547-4/06 $20.00 © 2006 IEEE
IHP GmbH - Innovations for High Perform-
ance Microelectronics
{Ebert, Dietterle}@ihp-microelectronics.com
biomedical clothes that combine sensing with on-body
processing for making diagnoses, detecting trends, and
recommending reactive actions.
2. BASUMA project
The BASUMA [3] consortium brings together aca-
demic and industrial expertise from electronics, com-
munications, bioengineering, and medicine in order to
develop a distributed wireless body sensor platform for
continuous health monitoring of chronically ill patients
in their own homes. The BASUMA project particularly
focuses on improving the treatment of chronic obstruc-
tive pulmonary disease (COPD) patients and the ambu-
latory chemo therapy of women suffering from breast
cancer. These diseases have been chosen because of the
great economic and/or social burden they generate in
western countries.
By means of COPD we explain below how
BASUMA improves the quality-of-life of COPD pa-
tients while reducing COPD-related costs.
Table 1 depicts the annual costs caused by COPD in
Germany and the potential savings by BASUMA. As-
suming conservatively a 3.3% prevalence of COPD,
there are a total of 2.7 million patients with COPD in
Germany, which can be classified into different degrees
of severity. The annual COPD-related costs per patient
are ranging from € 2,364 (mild) to € 6,585 (severe) [4],
resulting in a total of M€ 8,191 spent for COPD per
year in Germany. BASUMA pursues the aim to reduce
costs by preventing exacerbations, decreasing the days
of incapacity for work, and decelerating the aggrava-
tion of COPD. With BASUMA we suppose to accom-
plish 50 per cent less hospitalizations due to exacerba-
tions (which translates to a saving of € 2,190 per omit-
ted inpatient treatment of an exacerbation [4]), 50%
less absence from work, and a deceleration of aggrava-
tion by 25 per cent. Thereby BASUMA has the poten-
tial to cut the COPD-related costs in Germany down by
25%, leading to a total saving of M€ 2,047.
 Table 1. Potential COPD costs savings by BASUMA extreme falls and thereby sustains the state of health on
a higher level as shown in Figure 2.
COPD in Level of severity
Total
Germany mild moderate severe
Patients 1,786 649 271 2,706
in 000s (66%) (24%) (10%) (100%)
Annual costs of COPD in Germany
Per € 3,027
€ 2,364 € 3,332 € 6,585
patient (average)
Total M€ 4,222 M€ 2,164 M€ 1,782 M€ 8,191
Potential annual COPD costs savings by BASUMA
Per € 757
€ 434 € 662 € 3,115
patient (average)
Total M€ 774 M€ 430 M€ 843 M€ 2,047
Figure 2. BASUMA sustains a higher state of health
3. Approach
Today patients perceive the gradual aggravation of 4. Body sensor network
their state of health not until it has evolved into an al-
ready critical condition. As a result counteractive Continuous patient monitoring requires that unob-
measures (such as taking a drug or seeing a doctor) are trusive tiny medical sensors measuring and analyzing
taken too late. The BASUMA approach is to mitigate for example electrocardiogram, pulse oximetry, and
the discrepancy between the objective patient’s state of lung sound are placed on the patient’s body and form a
health and how it is subjectively perceived by the pa- wireless body sensor network as illustrated in Figure 3.
tient as illustrated in Figure 1. The medical data can be enriched with context infor-
mation by attaching further wireless activity sensors
(e.g., accelerometers and gyro meters) to the subject. If
required the BASUMA system may prompt the patient
to perform an additional spot check by using a hand-
held device such as a spirometer, which becomes also
part of the body sensor network. A mobile phone or a
PDA acting as gateway to global backbone networks
(e.g., the Internet) completes the system by providing
connectivity to a remote medical center.

Body-worn sensors Handheld sensors

Lung Spirometry
Figure 1. BASUMA timely detects critical conditions sound
ROS
Continuous monitoring and analyzing the patient’s ECG
vital signs is the key for detecting when her state of
health changes to the worse. The BASUMA system Lactate
Blood
alerts and recommends actions (as predefined by the pressure
attending physician) to the patient in a timely manner Handhelds
before a critical condition arises, thereby effectively Impedance Mobile
preventing severe exacerbations, which are costly and -
phone
what is more- harmful to the patient. This empowers Pulse
patients to take more personal responsibility and to be oximetry
more proactive in general. Medical practitioners are
encouraged to draw up a therapy plan in close partner- Thermistor
ship with the patient, which in turn improves their com-
pliance. Compared to the traditional model where the Accelerometer
patient’s state of health is subject to major changes up
to and including emergency cases, BASUMA avoids Figure 3. BASUMA body sensor network

Proceedings of the International Workshop on Wearable and Implantable Body Sensor Networks (BSN’06)
0-7695-2547-4/06 $20.00 © 2006 IEEE
 We pursue the concept of a distributed system with the popular self-contained VHDL model of the
a wireless peer-to-peer network without central control- LEON2 processor. We chose the highly configur-
ler. As a consequence of shifting the intelligence to- able LEON2 model as the basis for our system-on-
wards the sensors, the body sensor network consists of chip design of the BASUMA core module.
smart, self-contained, wireless sensors that communi-  Radio transceiver
cate with one another. Because peer-to-peer networks The BASUMA module exploits ultra wideband
are not dependent on any particular component, they (UWB) as wireless communication technology.
are failure-tolerant, i.e. even if one component fails the Since UWB uses a very broadband signal with ex-
remaining parts of the system continue to operate. Fur- treme low power spectral density, a highly energy-
thermore this approach is preferable when sensors need efficient and robust communication is assured. We
to communicate with each other. A non-invasive con- are designing a UWB analog front-end and a cor-
tinuous cuff-less blood pressure sensor is a good exam- responding digital baseband processor providing a
ple for this: by combining the signals from at least one data rate in the order of 20 Mbit/s, which allows
distal pulse wave sensor and a single lead ECG, the saving energy by sending data in short bursts.
pulse wave velocity can be computed, which shows a  Embedded operating system
linear relationship with the blood pressure. Reflex (Realtime Event FLow EXecutive) is a ge-
neric, preemptive, and event-driven operating sys-
5. Architecture tem developed at the BTU Cottbus for embedded
systems that require small memory footprint, ro-
The BASUMA system architecture serving one or bustness, and real-time capabilities [6]. We already
more medical applications is depicted in Figure 4. The ported the Reflex implementation to the LEON2
pillars represent the wireless devices, namely the body processor, thereby validating that Reflex requires
sensors such as an ECG or an SpO2 sensor, the hand- only a footprint of 4 KB of RAM.
held measurements such as an ROS (reactive oxygen  Wireless communication protocols
species) sensor and a spirometer, and the gateways to The BASUMA medium access control protocol is
other networks such as a mobile phone or a PDA. based on the IEEE 802.15.3 standard, which sup-
ports networks consisting of a coordinator and up
to 255 associated devices, beacons for synchroni-
… zation, guaranteed time slots for ensuring quality
Medical Heart
…
Hearttransplantation
transplantation of service, energy-efficient synchronous and asyn-
Apps Breast
Breastcancer
cancer
chronous sleep modes, 128-bit AES encryption,
COPD
COPD
adaptation of transmission power, and handover to
ECG SpO2 Lung
Sensor Sensor Sound
ROS
Sensor
Mobile
Phone
… an alternative coordinator in case of failure.
Middleware
Monitor
 Middleware
SW Device & Service discovery + Security + … The BASUMA middleware allows the wireless
Wireless Communication Protocols nodes to operate as a distributed system. It facili-
IEEE 802.15.3 tates users to easily form a wireless body sensor
Embedded Operating System network and applications to discover the presence
Reflex OS of other nodes and their services on the network, to
Processor Radio invoke services on other nodes, and to subscribe to
HW LEON2 Ultra Wideband (UWB)
events. In addition, the middleware provides sup-
port for security and gateway functionality for
Figure 4. BASUMA system architecture bridging the body sensor network to other net-
works such as an IEEE 802.11-based wireless
Each wireless device contains in addition to device- LAN or a GSM/UMTS-based wide area network.
specific components (such as ECG electrodes and pre-  Interfaces
amplifier) a generic BASUMA module, which consists Standard interfaces, for example UART, SPI, or
of the following hardware and software elements: GPIO, are provided for connecting medical sen-
sors to the BASUMA platform.
 Processor
The heart of the BASUMA module is the LEON2 Concerning the wireless transmission part we work
32-bit SPARC V8 processor core with several on- also on an alternative solution offering even higher data
chip peripheral units [5]. Gaisler Research main- rates and more flexibility by exploiting OFDM and a
tains and distributes under the GNU LGPL license distributed multiple access mechanism [7].

Proceedings of the International Workshop on Wearable and Implantable Body Sensor Networks (BSN’06)
0-7695-2547-4/06 $20.00 © 2006 IEEE
 6. System-on-chip
To reduce the overall power consumption, we pur-
sue a system-on-chip (SoC) integration of all the com-
ponents that constitute the BASUMA core module. The
BASUMA SoC (see Figure 5) contains all components
as described before and additionally dedicated power
management, memory and timer modules. The compo-
nents are interconnected with the ultra-fast AMBA
(AHB) bus system. For the SoC implementation and
fabrication we use IHP’s 0.25 μm SiGe:C-BiCMOS
technology and fab facilities.
Figure 5. BASUMA system-on-chip schemata
7. Status
To validate the envisaged medical use cases and
middleware concepts, we realized a first BASUMA
demonstrator based on the Philips AquisGrain wireless
sensor platform compliant to IEEE 802.15.4. Several
sensors such as ECG, SpO2, blood pressure, gyro meter
and accelerometer have been integrated to AquisGrain
including the required communication protocols for
transmitting the medical waveforms. We developed an
AquisGrain-enabled mobile phone and a PDA acting as
gateway to GSM, WLAN, and Bluetooth networks.
Initial functional demonstrators of novel sensors that
are developed within the BASUMA project (i.e.
chemical biosensors for measuring reactive oxygen
species (ROS) and lactate in respiratory air, and a lung
sound monitor) are available.
An implementation of the LEON2 processor core as
FPGA board has been completed in 2005, which serves
as an experimental platform for developing and testing
the digital baseband processor and the BASUMA MAC
software. Furthermore the first transceiver chips have
been produced and tested. A first demonstration of the
UWB- and IEEE 802.15.3-based wireless transceiver is
planned for mid 2006, while initial samples of the
complete BASUMA SoC are scheduled for the end of
2006.
Proceedings of the International Workshop on Wearable and Implantable Body Sensor Networks (BSN’06)
0-7695-2547-4/06 $20.00 © 2006 IEEE
Clinical trials for validating the BASUMA system
are under way.
8. Conclusions
We presented the BASUMA approach of continu-
ously monitoring and analyzing the state of health to
prevent arising critical conditions. This improves qual-
ity-of-life of chronically ill patients by sustaining their
state of health on a higher level while at the same time
it reduces health expenditures. BASUMA empowers
patients to take more personal responsibility and im-
proves their compliance.
BASUMA is developing a robust and energy-
efficient platform based on UWB and IEEE 802.15.3
as system-on-chip for wireless body sensor networks,
enabling continuous health monitoring of chronically ill
patients in homecare settings.
Trials with COPD and chemo therapy patients for
validating the BASUMA system are under way.
9. Acknowledgements
We would like to thank all of our BASUMA project
colleagues from: ABS GmbH, AOK Berlin-
Brandenburg, BST Biosensor GmbH, Charité Univer-
sitätsmedizin Berlin, Filt GmbH, TES GmbH, TU
Hamburg-Harburg, and University of Potsdam.
We would also like to express our gratitude to the
Bundesministerium für Wirtschaft und Arbeit for par-
tially funding this work under contract 01MT0305.
10. References
[1] HealthService24 Project,
http://www.healthservice24.com/
[2] MyHeart Project,
http://www.hitech-projects.com/euprojects/myheart/
[3] BASUMA Project, http://www.basuma.de/
[4] D. Nowak, E.S. Dietrich, P. Oberender, K. Überla, U.
Reitberger, C. Schlegel, et al., “Cost-of-illness study for the
costs of COPD in Germany”, In: Pneumologie 58, Thieme,
Stuttgart, 2004, pp. 837-844
[5] Gaisler Research, “LEON2 Processor”, Datasheet,
http://www.gaisler.com/products/leon2/leon.html
[6] K. Walther, R. Hemmerling, J. Nolte, "Generic Trigger
Variables and Event Flow Wrappers in Reflex", In: Proc. of
ECOOP Workshop on Programming Languages and Operat-
ing Systems, Oslo, Norway, 2004
[7] N. Toender, M. Reinert, L. Ohliger, H. Rohling, „Flexi-
ble OFDM Demonstrator with a Single FPGA Implementa-
tion”, In: Proc. of 9th International OFDM-Workshop, Ham-
burg, Germany, 2004