Analysis of EEG to quantify depth of anesthesia using hidden markov model

Abstract

Real-time quantification of the patient’s consciousness level during anesthesia is an important issue to avoid intraoperative awareness and post-operative side effects. A depth-of-anesthesia (DoA) monitoring method called Bispectral Index (BIS) is generally used for this purpose. However, BIS is known to be inaccu-rate at the transitory state, and also shows a critical time delay in quantifying the patient’s consciousness level.This thesis introduces a novel method to reduce the response time in the quantification process. This thesis develops a new index called HDoA by analyzing EEG using Hidden Markov Model. The proposed ap-proach is composed by two steps, training and testing. In the training step, two HMM, awakened and anesthetized model are learned based on the each training set. In the testing step, by evaluating the probability of producing the testing EEG from two models respectively, the index HDoA is derived. Since the evaluation of DoA using HMM is training based method, it have better performance with more training process. In addition, the run-time complexity of proposed method is lower than other training-testing methods.Experiments show that HDoA has a high correlation with BIS at a steady state, and outperforms BIS in two ways: (1) higher Fisher Score, (2) shorter delay time in transition state, and (3) stability in poor signal quality. The validity of HDoA has been tested by 64 real clinical data.