Evaluation of Chaos on Electroencephalogram in Different Depths of Anesthesia

Today having monitors and instruments which are able to automatically and precisely determine the depth of anesthesia from the electroencephalogram (EEG) signal is important. The purpose of this was is to provide an approach to assess the dynamics of brain chaos and its electrical activity in order to take advantage of the achievements of this theory in cognitive science.
According to the chaos theorem, the chaotic features of the electroencephalogram signal in different anesthesia levels have been extracted and evaluated as a chaotic system trajectories. In order to evaluate the effect of anesthesia level on the chaotic behavior of electroencephalogram signal, different models created based on the random forest, and the support vector machine modeling. We proposed a procedure to extract largest Lyapunov exponential and Higuchi’s fractal dimension as chaotic features from one channel electroencephalogram in 20 patients under the different depths of anesthesia with sevoflurane; the evaluation was done using K-fold procedure.
Findings: Evaluation of extracted models indicated that mentioned models had repeatability and separability with the accuracy of more than 93%.
Results show that the brain and its electrical activities have chaotic dynamism. Therefore, we can take advantage of chaos theorem in developing of anesthesia monitoring, as well as in many other researches related to the cognitive sciences by analyzing the electroencephalogram signal based on the chaos theorem.