Depth of anesthesia estimation by adaptive-network-based fuzzy inference system

One effective way to estimate the depth of anesthesia (DOA) from EEG is proposed. The scheme applies an adaptive-network-based fuzzy inference system (ANFIS) to integrate the extracted EEG characteristics such as complexity measure, approximate entropy, and spectral edge frequency for decision-making. The system was trained and tested using EEG data collected from three dog experiments under propofol anesthesia. The accuracy of the system attains 89.5%. Comparison with artificial neural networks was made.