You are here

TANIA

Title: Decision Support System for Control in Anesthesia

News: TANIA featured in ATHENA

Updates: TANIA Wiki (Restricted Access)

Partners:

  • Machine Learning Group, ULB, Bruxelles
  • TCTS Lab, FPMs, Mons
  • Department of Anesthesiology, Erasme Hospital, ULB, Bruxelles
  • Department of Anesthesiology, St. Luc Hospital, UCL, Louvain-la-Neuve 
  • Department of Anesthesiology, University Hospital , ULg
 

Funding: Waleo II, Walloon Region

Duration: 2006-2010 

Project outline and goals: 

A patient undergoing general anesthesia reacts dynamically to surgical stimuli of variable intensity. Consequently, the anesthetist adapts and modulates the concentration of the anesthetic agents with fast and short actions. The recorded data of the monitors of anesthesia corresponding to the information introduced by the anesthetist generate volumes of unexploited numerical data. Today, the anesthetist bases himself on his own calculations and memory, prior experiences, learned procedures and experiments to establish a plan of administration of the agents of anesthesia depending on the phases of the operation, surgical events, and the general state of the patient.
There exist data-mining techniques which are proving useful in fields such as bioinformatics, finance and marketing, search engines, etc, which could today aid the above described practice of the anesthetist by means of a decision support system based on models of behavior learned automatically starting from the data collected in the operating theatre suite.
The main objective of the project is to aid the anesthesiologists and anesthetists in using a computerized decision-making data-processing system which proposes sequences of administration of anesthetics and other drugs according to the information available from the patient’s general state, response to the surgery and the various phases of the operation. All the proposals for actions on anesthetics and other drugs will be posted on a monitoring system accessible to the expert anesthesiologist who will be able to validate them and possibly accept.
Personalized and customized, such a system will be based on an intelligent information system which works through machine learning algorithms starting from the training database collected as a preliminary in the departments of anesthesiology from the university hospitals which are partners of the project. We envisage two procedures for the proposed decision support system: first, a standard method where the system proposes sequences of actions in accordance with the intelligent information system and an advanced method of update where new data collected during the anesthesia will be added to the existing database in order to update the intelligence attained by the system. The first procedure could make available the service of the decision support system of anesthesia to all the practitioners in all conditions (for example, urgency during the night). The second method would be used to make the tool in conformity with the requirements of an informed user (for example an experienced anesthetist) who could be interested in better gauging the data-processing tool with his style of anesthesia.
We believe that such a computerized decision-making system would facilitate improved quality of the anesthesia offered to the patient while allowing therapeutic freedom to the doctor.

MLG researchers involved:

 

Theme by Danetsoft and Danang Probo Sayekti inspired by Maksimer