Share Email Print

Proceedings Paper

Parallel computer processing and modeling: applications for the ICU
Author(s): Grant Baxter; L. Alex Pranger; Nicole Draghic; Nathaniel M. Sims; William P. Wiesmann
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Current patient monitoring procedures in hospital intensive care units (ICUs) generate vast quantities of medical data, much of which is considered extemporaneous and not evaluated. Although sophisticated monitors to analyze individual types of patient data are routinely used in the hospital setting, this equipment lacks high order signal analysis tools for detecting long-term trends and correlations between different signals within a patient data set. Without the ability to continuously analyze disjoint sets of patient data, it is difficult to detect slow-forming complications. As a result, the early onset of conditions such as pneumonia or sepsis may not be apparent until the advanced stages. We report here on the development of a distributed software architecture test bed and software medical models to analyze both asynchronous and continuous patient data in real time. Hardware and software has been developed to support a multi-node distributed computer cluster capable of amassing data from multiple patient monitors and projecting near and long-term outcomes based upon the application of physiologic models to the incoming patient data stream. One computer acts as a central coordinating node; additional computers accommodate processing needs. A simple, non-clinical model for sepsis detection was implemented on the system for demonstration purposes. This work shows exceptional promise as a highly effective means to rapidly predict and thereby mitigate the effect of nosocomial infections.

Paper Details

Date Published: 22 July 2003
PDF: 8 pages
Proc. SPIE 4958, Advanced Biomedical and Clinical Diagnostic Systems, (22 July 2003); doi: 10.1117/12.477868
Show Author Affiliations
Grant Baxter, BioSTAR, Inc. (United States)
L. Alex Pranger, BioSTAR, Inc. (United States)
Nicole Draghic, BioSTAR, Inc. (United States)
Nathaniel M. Sims, Ctr. for Integration of Medicine and Innovative Technology (United States)
William P. Wiesmann, BioSTAR, Inc. (United States)

Published in SPIE Proceedings Vol. 4958:
Advanced Biomedical and Clinical Diagnostic Systems
Tuan Vo-Dinh; Warren S. Grundfest; David A. Benaron; Gerald E. Cohn, Editor(s)

© SPIE. Terms of Use
Back to Top