Wireless body area networks for ambulatory health monitoring

Transcription

1 Wireless body area networks for ambulatory health monitoring with prototype demonstration August 25, 2006 Chris A. Otto Electrical and Computer Engineering Department University of Alabama in Huntsville

2 Outline Introduction Proposed Solution WBAN Prototype WBAN Wireless Communications Embedded Software Personal Server Application Demonstration Conclusions

3 Motivation Healthcare is the Largest Segment of US Economy $1.8 Trillion in 2004 (15% of GDP) $200 Million Informal Care Givers $4178 per capita (50% more than the next nearest nation) - US is less than 10th in life expectancy - US is 26th in infant mortality rates Pending Crisis Retiring Baby Boomers Elderly is the Largest Growing Age Group 45 million Uninsured

4 Motivation Current Healthcare Systems are Centralized, Focused on Reacting to Illness We are in need of Distributed Systems, Focused on Proactive Wellness Management Healthcare Spending by Category Drugs 15% Diagnostic 4% Causes of Death in the US in 2000 Tobacco 18.1% Poor diet and physical inactivity 16.0% Alcohol consumption 3.5% Microbial agents 3.1% Toxic agents 2.3% Motor vehicles 1.8% Firearms 1.2% Treatments, Hospital Stays, and Rehabilitation 81%

5 Ambulatory Health Monitoring Wearable Systems For Health Monitoring Close monitoring of Vital Signs Quantitative Feedback Computer Assisted Rehabilitation Holter Monitors Data Recorders ( < 24 hours) Post-session Analysis Telemedicine Systems

6 CardioLabs Ambulatory Heart Monitoring Event-based Recorders Loop Recorders (32 min) Patient Presses event button during episode Data Extracted for Post-Analysis

7 CardioNet Mobile Cardiac Outpatient Telemetry (MCOT) Wireless Sensor and Wireless Monitor Heartbeat by Heartbeat Monitoring Arrythmia event detection

8 Heart Rate Monitors Polar Electro, Suunto, Timex, Reebok, Integration into Fitness Equipment Real-time Heart rate Some Journal capabilities

9 ActiWatch Cambridge Neurotechnology Actigraphy Single Axis Accelerometer Clinical Research Sleep / Wake Patterns Sleep Disorders Periodic Leg Movement (PLMS) Infant Monitoring

10 Body Media (bodybugg) Multi-modal sensing Single axis-accelerometer (motion) (2) Temperature Sensors (heat flux) Galvanic Skin Response (GSR) Upload Data using USB Calorie Consumption Estimation Proprietary Algorithms Clinically tested

11 CodeBlue Harvard University, Boston University School of Management, and Boston Medical Center, 10Blade (start-up) Real-time Triage, Disaster Relief Pulse Oximeters, ECG, accelerometers

12 Outline Introduction Proposed Solution WBAN Prototype WBAN Wireless Communications Embedded Software Personal Server Application Demonstration Conclusions

13 Proposed Solution Wireless Body Area Network (WBAN) for Ambulatory Health Monitoring Mobility Increased Quality of Life Multimodal Physiological Monitoring Hierarchical Multi-tier Telemedicine System

14 System Architecture A WBAN E ZigBee or Bluetooth WWAN WLAN User User 1 2 WWAN (GPRS) User N Tier 3: MS A User 2 User 1 Tier 1: WBAN A User N WLAN (Wi-Fi) nc Tier 2: PS Internet Medical Server Healthcare provider Emergency Informal caregiver

15 Outline Introduction Proposed Solution WBAN Prototype WBAN Wireless Communications Embedded Software Personal Server Application Demonstration Conclusions

16 WBAN Prototype - Goals Implement a Working WBAN Activity Monitor / Motion Sensor ECG Sensor Network Coordinator Personal Server Explore Challenges Sensor Fusion On-Sensor Processing Communications Power Efficiency

17 Hardware Architecture - Tmote Sky CC2420 Radio 2.4 GHz IEEE Compliant CS SPI ST Flash 1MB SPI CS Temperature Sensor I/O P4.2 3 SPI[0] P4.4 UART[0] 2 I 2 C[0] 2 GPIO 2 10-pin header Humidity Sensor Light Sensor JTAG 8-pin IDC header ADC[4] ADC[5] JTAG 7 TI MSP430 Microcontroller UART[1] ADC[0-3] 2 ADC[6-7] 2 GPIO 2 UserINT Reset 6-pin header 2 JTAG RX/TX FT232MB USB controller

18 Hardware Architecture - ISPM MSP430F KHz (ACLK) / ~4.6MHz DCO Clock (MCLK,SMCLK) 8KB Flash / 256B RAM Two (2) ADXL202 Accelerometers Two Analog ECG Channels 10-pin Tmote Sky Header ISPM Tmote TelosSky ADXL202 Accelerometer ADXL202 Accelerometer TI MSP430F1232 TI MSP430F149 CC2420 (ZigBee) Flash ECG Signal Conditioning USB Interface ECG electrodes

19 ActiS Sensor Nodes z 3 axis motion detection Step detection and gait analysis Activity induced Energy Estimation (AEE) y x

20 Intelligent Signal Processing Modules (ISPM) Polar Heart Rate Monitor ISPM for Activity / ECG Single chip 3-axis Accelerometer

21 Outline Introduction Proposed Solution WBAN Prototype WBAN Wireless Communications Embedded Software Personal Server Application Demonstration Conclusions

22 IEEE and ZigBee IEEE LR-WPAN Access 250 Kbps CSMA-CA ZigBee Network Application and Application Sublayer Security Star Network Topology

23 Power Efficient TDMA >90% Sensor Power from Radio Significant Power Savings From Disabling Radio Timeslots for Communication Distributed Events Concentrated Bursts Allows Radio to be disabled Extended Battery Life / Lower Weight Super Cycle Beacon Slot#1 Slot#2 Slot#3 Beacon Listen Tr. Listen 50 ms 100 ms 150 ms 1000 ms Time

24 TDMA Means Low Power Deterministic RF timeslots Radio can be disabled Extend battery life 25 I [ma] Listen Transmit Idle Time [sec] 20 I [ma] superframe Time [sec]

25 Battery Life Actis Processed ACC Signals 40 Days Actis Raw ACC Signals 20 2xAA Alkaline batteries (50gr) Li-Ion battery (22gr) Li-Ion ipod mini battery (6gr) Channel utilization [% ]

26 Typical Message Flow Personal Server eactis (ECG) Sensor NC ACTIS_CALMSG Calibration Time ACK TIME ACTIS_EVENT_MASKMSG event = HRV, R Peak ACK ACTIS_CFG_STARTMSG ACK ACTIS_EVENT_MSG event = R Peak Session Started Heartbeat (R peak) Detected

27 Outline Introduction Proposed Solution WBAN Prototype WBAN Wireless Communications Embedded Software Personal Server Application Demonstration Conclusions

28 Embedded Software - TinyOS Lightweight Open Source Configuration Defined at Compile Time nesc (for Networked Embedded Sensors) Component model Task support Split-phase Operation (command / event model)

29 Network Coordinator WBAN Access for Personal Server Network Channel Management Distribute Global Timing

30 Network Coordinator

31 Time Synchronization - Motivation TDMA Efficient Sharing of Communication Channels Timeslot Assignments Beacon Prediction (Maximize Radio off) Correlating Intra-WBAN events Relative timing is important Synchronizing start time

32 Time Synchronization Protocol Flooding Time Synchronization Protocol (FTSP) MAC Layer Time Stamping Skew Compensation with Linear Regression SFD Capture Timer Process Send Beacon Transmission Beacon Reception Preamble SFD Length MAC Protocol Data Propagation Preamble SFD Length MAC Protocol Data Timestamp Timestamp SFD Capture Timer TimeSyncM (FTSP Algorithm)

33 Sensor Nodes Collect Physiological Data Process Data (Feature Extraction) Event Management Resource Management WBAN Communications

34 Sensor Nodes

35 Feature Extraction Step Detection Accelerometer based step recognition Step detected Vy[m/s] Alpha[rad] accy[g] Step Foot switch Time[s]

36 Feature Extraction Activity Estimation AEE t t = t δ AC( a x ) 2 + AC( a y ) 2 + AC( a z ) 2 dt Acc Signal HP Filter AC Signal Energy Estimator - + DC Signal Orientation

37 Feature Extraction AEE and Heart Rate Sitting Standing Slow walking Fast walking Running Sitting AEE 10 x Heart rate Time [min] Time [min]

38 Event Management AEE HR WBAN Pkts Out of range Burst uploads Time [min]

39 Event Management x 104 accx accy Motion accz Sensor 1 (TS2) ECG 3000 Sensor (TS1) Heart Beat Heart Beat Step Heart Beat Step Beacon Message Event Message with Timestamp Beacon Message NC TS1 TS2 TS3 NC TS1 TS2 TS3 Frame i-1 Frame i

40 Resource Management USART0 Scheduled Beacon Reserved For Flash Scheduled Timeslot Radio UART USART0 SPI SPI SPI SPI SPI SPI Collect Sample Collect Sample Collect Sample Collect Sample Collect Sample Collect Sample Rx Radio Flash Ops Tx Radio Tx Radio Tx Radio Tx Radio Time 25ms 50 ms 100 ms 150 ms

41 Outline Introduction Proposed Solution WBAN Prototype WBAN Wireless Communications Embedded Software Personal Server Application Demonstration Conclusions

42 Personal Health Server Sensor Node Identification and Configuration Sensor Fusion Session File Management Graphical User Interface (GUI)

43 Health Server Configuration

44 Health Server Real-time Display

45 Outline Introduction Proposed Solution WBAN Prototype WBAN Wireless Communications Embedded Software Personal Server Application Demonstration Conclusions

46 Outline Introduction Proposed Solution WBAN Prototype WBAN Wireless Communications Embedded Software Personal Server Application Demonstration Conclusions

47 Conclusions Working WBAN Prototype Custom-designed ISPM daughter cards Off-the-shelf wireless sensor platforms Standard IEEE communications Original Solutions Health monitoring specific, Power Efficient TDMA Scheme WBAN communication protocol On-sensor real-time signal processing Promising Technology Optimal Treatment of Disease Rehabilitation Low Cost Early Diagnosis Encourages Wellness Management Improved Mobility Low Weight (Increased Compliance)

48 Future Research Validation Security Hardware encryption of wireless communications Standard security mechanisms from the personal server to the upper levels of hierarchy Signal Processing Discern category of activity Improved Step Detection Improve ECG Analysis Upper Tier Development Integration into Electronic Medical Records (EMR)

49 Acknowledgements Dr. Emil Jovanov, Dr. Alex Milenkovic Elise Haley, Corey Sanders, Reggie McMurtrey, John Gober University of Alabama in Huntsville