Docstoc

Safe Composition of Complex Medical System - Dashboard

Document Sample
Safe Composition of Complex Medical System - Dashboard Powered By Docstoc
					Safe Composition of Complex
      Medical System

 University of Illinois at Urbana-Champaign
        Computer Science Department
                  Fall 2010
Medical Accidents
 More than 100,000 safety-related medical incidents a year in
  US
   • According to Health Grades


 Caused by human error
   • Drug-based adverse events
   • Overlooked device interlocks
   • Alarm fatigue


 Preventable by automated supervision
The MDPnP Program
 MDPnP : The Medical Device “Plug-and-Play”
   Chair: Dr. Julian Goldman
   Inter-disciplinary, multi-institutional medical device informatics
    research program.
   Leading patient-centric Integrated Clinical Environment (ICE)
    standard

 MDPnP group is based at CIMIT and Massachusetts General
  Hospital.
System Overview                                                        Clinician



                             Pre-condition
                               Checker



                        ICE Supervisor
                   Safety Interlock             False Alarm
                       Manager                   Detector

             Virtual
             Device
                           Virtual
                           Device       ....              Virtual
                                                          Device




    Device             Device            ....                 Device



                                      Patient
UIUC Demos
 We demonstrate three related safe MDPnP technologies:
   NASS Framework- Airway Laser Surgery
     Ensuring distributed medical devices to enter failure safe states when the
       supervisor loop fails
   Context Aware Alarm Processing
     Using correlations between vital signs under a given medical procedure to
      validate alarms or reconstruct missing measurement
   Patient Screening for Safe MDPnP
     Identify potential safety hazards
UIUC Demos
 The presentations are based on the live demo at
  MDPnP lab at CIMIT.

 We would like to thank all the participants from
  Mass General, FDA, UPenn and U of Kansas, for
  their comments.
Publications
 Y. Li,Y. Sun, P. Sondhi, L. Sha, and C. Zhai, “Reconstructing missing signals in
   multi-parameter physiologic data by mining the aligned contextual information”,
   in Proceedings of Computing in Cardiology Conference 2010, Belfast, Northern Ireland,
   United Kingdom, Sep. 26-29, 2010.

 P. Sondhi, J. Sun, C. Zhai, R. Sorrentino, M. S. Kohn, S. Ebadollahi, and Y. Li,
   “Medical case-based retrieval by leveraging medical ontology and physician
   feedback”, UIUC-IBM at ImageCLEF 2010. CLEF

 Y. Li, J. H, C. Zhai, and Y. Chen, “Improving one-class collaborative filtering by
   incorporating rich user information”, in CIKM 2010: Proceedings of the 19th ACM
   international conference on Information and knowledge management,Toronto, Canada,
   2010. ACM. (acceptance ratio: 13.4%)
Publications
 P. Sondhi, M. Gupta, C. Zhai and J. Hockenmaier, “Shallow information
  extraction from medical forum data”, in Proceedings of COLING 2010,pp. 1158-
  1166

 C. Kim, M. Sun, S. Mohan, H.Yun, L. Sha, and T. F. Abdelzaher, “ A
  framework for the safe interoperability of medical devices in the presence of
  network failures”, in Proceedings of the 1st ACM/IEEE International Conference of
  Cyber-Physical Systems, 2010, pp. 149-158

 M. Rahmaniheris, C. Kim, S. Bak, and L. Sha, “A multi-layer dependency
  framework for analysis of safety-critical embedded systems”, submitted to DSN
  2011

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:43
posted:11/21/2012
language:English
pages:8