Comparative clustering analysis of bispectral index series of brain activity

Bispectral index scale (BIS) is a continuous processed electroencephalogram (EEG) parameter that correlates to the patient’s level of brain activity, where 100 is awake and 0 (flat line) is dead. BIS was designed to correlate with “hypnotic” clinical endpoints (sedation, lack of awareness, and memory) and to track changes in the effects of anesthetics on the brain. In this study, an approach to utilize clustering methods is investigated in the analysis of BIS series data. Fuzzy c-Means (The FCM) and Fuzzy Neighborhood DBSCAN (FN-DBSCAN) algorithms are handled in the paper. The FN-DBSCAN algorithm is advantageous in such a way that it aggregates the speed of the well-known Density Based Spatial Clustering of Applications with Noise (DBSCAN) and the robustness of the Noise-Robust Fuzzy Joint Points (NRFJP) algorithms. As a result of the computational experiments, we can conclude that FN-DBSCAN method gives more realistic results to recognize the stable duration intervals and the BIS stages in the measurement series.