Development of FIR Models to Judge the Depth of Anesthesia during Surgical Operations

Introduction

The anaesthetic agents that are being administered during a surgical intervention control separately three different mechanisms in the patient. On the one hand, they control the pain, on the other, they control the muscle relaxation, and finally, they control the level of anaesthesia. All three mechanisms are equally important.

If not enough anaesthetics have been administered, it can happen that the patient suffers unbearable pain, but cannot express his problems, since he cannot move his muscles. It can also happen that the patient is not truly unconscious, experiences what is done to him, and consequently suffers a trauma, yet cannot communicate his problems, as his muscles do not react. Finally, it can happen that an unconscious and pain-free patient suddenly reacts reflexively to an intervention, because his muscles aren't sufficiently relaxed.

However, all anaesthetic agents are powerful poisons that invariably produce unwanted side effects. To this end, it is important that the smallest safe dosis of anaesthetics is administered.

On the one hand, anaesthesiologists determine the dosis of anaesthetic agents on the basis of static information, such as the gender, the age, and the weight of the patient; on the other hand, they make use of quantitative dynamic data, such as the heart rate, the respiration rate, and the blood pressure. Finally, the experienced anaesthesiologist relies on qualitative dynamic data, such as the skin color, the humidity of the skin, and the skin tonus.

Since experienced anaesthesiologists aren't available during all surgeries, efforts are under way that would automate the determination of the dosis of anaesthetic agents to be administered.

To this end, physiological signals of patients, such as the heart rate, the respiration rate, and the systolic blood pressure, were recorded in a hospital in Scotland during surgery, together with the decisions made by the anaesthesiologist concerning the amount of anaesthetic agents to be administered.

A FIR model of the anaesthesiologist (not the patient) was then identified that should be able to predict, how the anaesthesiologist would decide in a given situation.


Most Important Publications

  1. Nebot, A., F.E. Cellier, and D.A. Linkens (1996), Synthesis of an Anaesthetic Agent Administration System Using Fuzzy Inductive Reasoning, Artificial Intelligence in Medicine, 8(3), pp.147-166.

Sponsors


Deutsche Version
Homepage


Last modified: July 16, 2005 -- © François Cellier