Title :
Non-linear control for anaesthetic depth using neural networks and regression
Author :
Linkens, D.A. ; Rehman, H.U.
Author_Institution :
Dept. of Autom. Control & Syst. Eng., Sheffield Univ., UK
Abstract :
The control of depth of anesthesia using artificial neural networks (ANNs) is discussed. The backpropagation algorithm is used to train the network on surgical data. The same technique is used to simulate a model of a patient under the effect of an anesthetic agent. An alternate controller and a patient model (PM) are developed by means of regression analysis of surgical data. The ANN controller and the ANN-PM are studied under closed-loop conditions, and the results are compared with those obtained by regression
Keywords :
backpropagation; medical computing; neural nets; surgery; anaesthetic depth control; backpropagation; closed-loop conditions; medical computing; neural networks; nonlinear control; patient model; regression analysis; surgery; Artificial neural networks; Automatic control; Biological neural networks; Blood pressure; Expert systems; Heart rate; Neural networks; Open loop systems; Patient monitoring; Surgery;
Conference_Titel :
Intelligent Control, 1992., Proceedings of the 1992 IEEE International Symposium on
Conference_Location :
Glasgow
Print_ISBN :
0-7803-0546-9
DOI :
10.1109/ISIC.1992.225126
