An architecture for online comparison and validation of processing methods and computerized guidelines in intensive care units

  • Authors:

  • L. Allart;C. Vilhelm;H. Mehdaoui;H. Hubert;B. Sarrazin;D. Zitouni;M. Lemdani;P. Ravaux

  • Affiliations:

  • Faculty of Pharmacy and Biology of Lille, Department of Biomathematics, University of Lille 2, 3 rue du Professeur Laguesse, BP 83, 59006 Lille Cedex, France;Institute of Engineering in Health of Lille, University of Lille 2, 42 rue Ambroise Paré, 59120 Loos, France;Department of Intensive Care, Fort de France University Hospital, BP 632, 97261 Fort-de-France Cedex, France;Institute of Engineering in Health of Lille, University of Lille 2, 42 rue Ambroise Paré, 59120 Loos, France and Department of Public Health, EA 2694, University of Lille 2, France;Department of Intensive Care, Fort de France University Hospital, BP 632, 97261 Fort-de-France Cedex, France;Faculty of Pharmacy and Biology of Lille, Department of Biomathematics, University of Lille 2, 3 rue du Professeur Laguesse, BP 83, 59006 Lille Cedex, France;Faculty of Pharmacy and Biology of Lille, Department of Biomathematics, University of Lille 2, 3 rue du Professeur Laguesse, BP 83, 59006 Lille Cedex, France;Faculty of Pharmacy and Biology of Lille, Department of Biomathematics, University of Lille 2, 3 rue du Professeur Laguesse, BP 83, 59006 Lille Cedex, France

  • Venue:
  • Computer Methods and Programs in Biomedicine
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Clinical decision support systems are a combination of software techniques to help the clinicians in their medical decision making process via functionalities ranging from basic signal analysis to therapeutic planning and computerized guidelines. The algorithms providing all these functionalities must be very carefully validated on real patient data and must be confronted to everyday clinical practice. One of the main problems when developing these techniques is the difficulty to obtain high-quality complete patient records, comprising data coming both from the biomedical equipment (high-frequency signals), and from numerous other sources (therapeutics, imagery, clinical actions, etc.). In this paper, we present an infrastructure for developing and testing such software algorithms. It is based on a bedside workstation where testing different algorithms simultaneously on real-time data is possible in the ward. It is completed by a collaborative portal enabling different teams to test their software algorithms on the same patient records, making comparisons and cross-validations more easily.