A hierarchical system of on-line advisory for monitoring and controlling the depth of anaesthesia using self-organizing fuzzy logic

A hierarchical system has been developed to on-line advise on the concentration of inhaled volatile anaesthetics for controlling depth of anaesthesia. It merges on-line measurements (such as systolic arterial pressure and heart rate) and clinical information (such as sweating, lacrimation and movement) using a hierarchical architecture and self-organizing fuzzy logic for reasoning. It has been developed to predict depth of anaesthesia from either a ''hand-crafted'' anaesthetists' or machine-learning rule-base using self-organizing learning system and control the drug levels using self-organizing fuzzy logic algorithm. In this paper, machine-learning rule-base has been validated via tests with 10 patients off-line and 17 patients on-line. The drug controller rule-base has also been validated via pre-tuning on 10 off-line patients and testing on 17 on-line patients. After extensive validation of this system, this on-line approach has shown promise and very successful for reducing the recovery time in comparison with either 10 patients off-line or other research.