Depth of anesthesia estimation and control

A fully automated system was developed for the depth of anesthesia estimati on and control with the intravenous anesthetic, Propofol, The system determ ines the anesthesia depth by assessing the characteristics of the mid-laten cy auditory evoked potentials (MLAEP). The discrete time wavelet transforma tion was used for compacting the MLAEP which localizes the time and the fre quency of the waveform. Feature reduction utilizing step discriminant analy sis selected those wavelet coefficients which best distinguish the waveform s of those responders from the nonresponders. A total of four features chos en by such analysis coupled with the Propofol effect-site concentration wer e used to train a four-layer artificial neural network for classifying betw een the responders and the nonresponders. The Propofol is delivered by a me chanical syringe infusion pump controlled by Stanpump which also estimates the Propofol effect-site and plasma concentrations using a three-compartmen t pharmacokinetic model with the Tackley parameter set. In the animal exper iments on dogs, the system achieved a 89.2% accuracy rate for classifying a nesthesia depth. This result was further improved when running in real-time with a confidence level estimator which evaluates the reliability of each neural network output. The anesthesia level is adjusted by scheduled increm entation and a fuzzy-logic based controller which assesses the mean arteria l pressure and/or the heart rate for decrementation as necessary. Various s afety mechanisms are implemented to safeguard the patient from erratic cont roller actions caused by external disturbances, This system completed with a friendly interface has shown satisfactory performance in estimating and c ontrolling the depth of anesthesia.