Discrimination of anesthetic states using midlatency auditory evoked potentials and artificial neural networks

This study was undertaken to determine whether Artificial Neural Network (ANN) processing of Mid-latency Auditory Evoked Potentials (MLAEP) can identify different anesthetic states during propofol anesthesia, and to determine those parameters which are most useful in the identification process. Twenty one (21) patients undergoing elective laparotomy were studied. To maintain general anesthesia, the patients received propofol. Epidural analgesia at the level of T4-5 blocked painful stimuli. MLAEP was recorded continuously with patients awake, during induction, during maintenance of general anesthesia, and during emergence until the patients were recovered from anesthesia. Four-layer artificial neural networks (ANN) were used to model the relationship between the parameters of the MLAEP and the 4 different states (awake, adequate anesthesia, during/before intraoperative movement, and emergence from anesthesia). The identification accuracy is, respectively, as follows: 97.5%, 88.6%, 84.4%, 90.2%, by 5 latencies and 97.1%, 85.7%, 80.0%, 86.4%, by the combination of 5 latencies and 3 amplitudes. The MLAEP has enough information for identifying different states, especially in its latencies. A nonlinear discrimination approach, such as the ANN, can effectively capture the relation between the MLAEP patterns and the different states of anesthesia