Predicting the depth of anaesthesia with 40-Hz ASSR

Abstract

An algorithm for predicting the amplitude of 40-Hz auditory steady state response (ASSR), which is an indicative of depth of anaesthesia, is presented in this paper. The amplitude of 40-Hz ASSR is an indicative of the depth of anaesthesia. Predicting this amplitude will help anesthesiologists in choosing the dose of the anaesthetic agents and have an estimation of the patients' next depth of anaesthesia state. The method is applied to the 40-Hz ASSR signals recorded from 20 human subjects during surgical operation. The algorithm uses sequential feature selection and multi-linear regression for choosing the most informative features and estimating the amplitude of ASSR on the next cycle and at the end of induction stage, when the patient looses his/her eyelash reflex. The algorithm uses the dose of anaesthetic drug, some of the demographics and medical parameters of the patients and their previous ASSR cycles for prediction. As the system provided with more ASSR cycles, the error in prediction decreases and estimation gets more accurate. The method is applied on the induction phase where patients' depth of anaesthesia changes very fast but it is applicable to all other stages of anaesthesia as well.