ZipCare EndtoEnd Platforms for Personalized Continuous HealthCare

1
ZipCareEnd-to-End Platforms for Personalized
Continuous HealthCare
Prof. Rahul Mangharam Electrical Systems
Engineering University of Pennsylvania

• Feb 2008

2
What is ZipCare?
ZipCare Health-Care Sensor Networks

• Project Goal
• System Components
• Responsibilities
• Timeline
• Next Steps

3
Project Goal

• Design and prototype of a personalized on-body
  medical sensor network with end-to-end
  monitoring, control and data interpretation.

4
Use Cases

• Chronic Patient Care
• Baby Boomers (60 years) are willing to pay
  300-500 per month for remote health monitoring
• Drug Trial Patient Monitoring
• Monitor side-effects of cardio-active drugs such
  as Vioxx, Viagra and Celebrex
• Monitor Pacemaker Performance
• Monitor cardiac arrhythmias and pacemaker
  response
• Continuous Care Database
• Long-term patient data for quantative and machine
  diagnosis
• Competitive medical device testing on long-term
  medical data

5
myBSN Benefits

• Increases accuracy of clinical alarms
• Increases Nurse Productivity
• Earlier Patient Mobility out of the hospital
• Faster Patient Transport
• Makes existing monitors wireless

BSN Body Sensor Network
6
System Components

• On-body Sensor Network Hardware
• Low-cost and low-power wearable health-strip
• Adaptive Embedded Operating System
• Adaptive software for context-aware monitoring
  and alerts
• BSN Middleware and Network Management
• Medical data monitoring, summarization and
  reporting
• Remote control of on-body medical devices
• Medical Data mining and Interpretation
• Patient medical data management
• Stream processing for preliminary data
  interpretation

7
Hardware

• Focus on ECG most prominent chronic disease
• Battery-operated patch features low-power radio,
  microcontroller, ECG and body motion sensors
• On-body Wearable waterproof flex-circuit
• Cost

Early Prototype
Model of evantual flex-circuit
8
Embedded Software

• Real-Time Embedded Sensor Operating System
• Pre-emptive multi-tasking with full resource
  management
• Time synchronized networking for max node
  lifetime
• Adaptive Parametric and Programmable control
• On-demand task and network scheduling for
  end-to-end delay-bounded communication
• Dynamic Task Partitioning between on-body node,
  mobile gateway and remote server

Real-Time Sensor OS
Adaptive Virtual Machine
9
Middleware Network Management

• Network Abstraction and Macro-programming
• Data Monitoring Aggregation Policies
• Remote Node Control (Parametric Programmable)
• Network Management and Security

10
Medical Data Interpretation

• Context-aware ECG data interpretation
• Based on patient mobility, medication, medical
  history and environmental factors
• Cardiac Arrhythmia Algorithms
• For remote server and on-body embedded system
• Data-mining for Quantative Diagnosis
• Over different time-scales
• Sensor fusion

11
Responsibilities

• Embedded Hardware UPenn
• Adaptive Sensor RTOS UPenn
• Embedded Application Software Infosys UPenn
• BSN Middleware Infosys
• Medical Data Interpretation Infosys
• Prototyping and Initial Trials Infosys UPenn

SETLab-Infosys
mLab-UPenn
12
Project Timeline 18 months
- Coupled components
13
Project Outcomes

• End-to-End Healthcare Solution
• Expertise in Embedded RTOS VM design
• Real On-body Hardware Manufactured in Bangalore
• Ready to deploy solution for any on-body sensor

14
Next Steps

• Converge on Project Goals
• Technical discussion on System Design
• Set realistic milestones with deliverables
• Discuss group communication and visiting
  personnel
• Formalize collaboration