EEDS and CPGs

1

Information Systems Engineering IS Research
Group Rob Kelly Associate Chair, Computer
Science

2
IS Research Projects

• Participating Faculty
• Computer Science R. Kelly
• Stony Brook Health Science Center A. Bicker,A.
  Singer, M. LoGuidice
• Marine Science R. Wilson, S. Piontkovski
• In-Place Research Contract Support
• Emergency Medicine Foundation (through 12/2006)
• Millennium NYStar (through 2006)
• US HRSA (through 2007)
• Sea Grant, NY
• NSF Division of Environmental Biology (through
  2006)

3
Projects Themes

• Health Science Systems
• Process reengineering
• Reduction in medical errors
• Knowledge representation
• Marine Science Systems
• Sensor systems
• Web data display
• Image processing
• Wireless applications

4
US Healthcare Issues

• Costs are enormous
• Quality is not consistent with cost
• Error rate is significant
• Adverse events in approximately 3 of hospital
  stays
• More than 50,000 fatalities per year

5
Anesthesiology Record Keeper

• Support SB Hospital
• Problem
• Improve accuracy of anesthesia data recording
• Reduce documentation requirement for
  anesthesiologists
• Approach
• Analyze OR processes
• Automatic monitor data capture
• Integrate with hospital data
• Allow data overrides
• Status
• Now in use in all Stony Brook Hospital Operating
  Rooms (ORs)
• DB allows for case analysis and data mining

6
Monitor QA

• Support - NYStar/US HRSA
• Approach
• Record monitor data during surgery for a large
  data sample
• Develop techniques for analyzing and comparing
  these recorded data streams
• Build a flexible network of stream integration
  modules
• Build stream analyzers to identify potential
  harmful conditions that are otherwise
  unobservable
• Capture secondary patient outcomes
• Relate secondary outcomes to monitor patterns and
  drug administration
• Provide stream data analysis for clinical trial
  drug effectiveness studies

7
Emergency Room System

• Support Emergency Medicine Foundation
  NYStar/US HRSA
• Problem
• Warning indicators missed or received following
  discharge
• Patient candidates for clinical trials not
  identified
• Approach
• Include an automated rule base for discharge
  analysis and clinical trial candidates
• Automatically obtain patient clinical data
• Process electronic rules against patient clinical
  data
• Provide hospital alerts to previously
  unidentified conditions

8
Marine Data Display

• Problem
• Develop a real-time system for the capture,
  analysis, and display of water/air data
• Approach
• Provide a Web-based interface with images
  generated in response to user requests
• Integrate Google Maps data with marine charts and
  images
• Integrate historical marine data with real-time
  data

9
LI Sound Sensor Network

• Stony Brook has instrumented the P.T. Barnum
  vessel with air and water sensors
• Sensors connected to a shipboard server
• Transmission of sensor data to ground station
  approximately once per minute
• Additional sensors planned

10
LI Sound and Plankton Display

• Real-time for LI Sound
• Historical data for Europe and Asia
• Image combination
• Merge of chart and image data
• Data selection

11
Questions