More Related Content Similar to AMA-IEEE Presentation : a low-power and reliable body area network platform for rehabilitation applications (20) AMA-IEEE Presentation : a low-power and reliable body area network platform for rehabilitation applications1. March 15, 2010
A Low-power and Reliable Body Area Network
Platform for Rehabilitation Applications
Fabien Massé 1
Shyamal Patel 2,3, Julien Penders 1, Bert Gyselinckx 1, and Paolo Bonato 2,4
1 Holst Centre / imec, Eindhoven, The Netherlands
2 Dept. of Physical Medicine and Rehabilitation, Harvard Medical School, Boston MA
3 Dept. of Electrical and Computer Engineering, Northeastern University, Boston MA
4 Harvard-MIT Division of Health Sciences and Technology, Cambridge MA
First AMA-IEEE Conference
2. © Holst Centre < 2
Background and Motivation
• Current limitations in rehabilitation monitoring
In-patient: limited feedback to the patients on
his/her exercises EKG &
Respiration
Lack of tools for long-term quantitative
assessment
Out-patient: No effective way to get information
about functional gain in daily life [1]
• Key advantages of Body Area Networks
Wearable >> Comfort of use and set-up
Low-power >> Longitudinal assessment of
patient’s recovery
Motion
Reliable >> High data integrity for on-line and
off-line recordings
Real-time >> Feedback or close-loop systems
A Low-power and Reliable BAN Platform for Rehabilitation Applications
3. © Holst Centre < 3
Low-power and reliable BAN platform
• Low-power
Imec’s ultra-low-power sensors [2]
3+ days of autonomy while continuously
transmitting the data
Lightweight: <20 grams
Small form factor: 52 x 32 x 15 mm3
• Reliable communication
Optimized wireless communication based on quality
of service rules [3]
Reduce data losses while maintaining limited latency
• Main features
Multiple nodes: up to 10 in the same network
Multiple sensors: ECG, EMG, Respiration,
Acceleration
Tunable sampling frequencies: Up to 1KHz
Wireless transmission or data logging on the nodes
A Low-power and Reliable BAN Platform for Rehabilitation Applications
4. © Holst Centre < 4
Platform architecture
Ultra low-power biopotential sensor[1]
Processing unit (backside) •ECG, EMG, EEG signals
Texas Instrument MSP430F1611 •Ultra Low Power Dissipation 21 μA @ 3V
• 8MHz
• 10KB RAM/48KB ROM
Power
Optional Accelerometer management
Analog Devices ADXL330
• -/+ 3g Top | Bottom
Radio transceiver Data storage
Nordic Semi nRF24L01 SD-card support for
• 2.4GHz / 2Mbps accurate offline data analysis
• Communication protocol
Star network : TDMA-based MAC protocol
Application-oriented Quality-of-Service (QoS) layer [2]
Application-oriented retransmission mechanism
Efficient balancing of : data integrity, autonomy,
and latency Radio communication reliability
A Low-power and Reliable BAN Platform for Rehabilitation Applications
5. © Holst Centre < 5
Preliminary evaluation for in- and out- patients
• Validation protocol
Two ambulatory environments
Office environment > daily-working activities
Clinical setting > limited displacements 10
Sensor setup 6 nodes 9 7.97 %
Office environment
Clinical environment
All nodes : 3D Acceleration (40 Hz) 8
Two nodes : extra EMG (500Hz) 7
Packet Error Rate (%)
Central node : extra ECG (200Hz) 6 4.95 %
5
• Qualitative feedback from clinicians 4
Comfortable 3
1.29 % 1.35%
Easy-to-setup User-friendly GUI 2
0.80 %
0.27%
Multi-modal sensor network
1
0
Real-time application Balenced latency Offline
latency <300 ms latency <1000 ms No latency constraints
A Low-power and Reliable BAN Platform for Rehabilitation Applications
6. © Holst Centre < 6
Future work
• Explore ways to cope with bursts of packet losses
Local processing
Context-aware QoS layer
On-node data storage
• Clinical trials
On epileptic patients in clinical environment
Scheduled for summer 2010
References
[1] Bonato P., “Wearable sensors/systems and their impact on biomedical engineering”, IEEE EMBS
Magazine 2003
[2] Yazicioglu R.F. et al., “A 60 μW 60 nV/√Hz readout front-end for portable biopotential acquisition
systems”, IEEE ISSC Conf, 2006
[3] Massé F. and Penders J., “Quality-of-Service in BAN: PER reduction and its trade-offs”, BSN 2010
A Low-power and Reliable BAN Platform for Rehabilitation Applications