Patients’ EEG analysis based on multi-indicator dynamic analysis measure for supporting brain death determination

Abstract

It is crucial to analyze coma and quasi-brain-death patients’ EEG (electroencephalography) by using different signal processing methods, in order to provide reliable scientific references for supporting BDD (brain death determination). In this paper, we proposed the multi-indicator dynamic analysis measure which was by combining Dynamic 2T-EMD (turning tangent empirical mode decomposition) and Dynamic ApEn (approximate entropy) to comprehensively analyze offline coma and quasi-brain-death patients’ EEG from dynamic EEG energy and dynamic complexity. Firstly, 60s EEG randomly selected from 36 cases of patients’ EEG (coma: 19; quasi-brain-death: 17) were analyzed to show the overall dynamic energy and complexity distribution for 2 groups. Secondly, one coma patient’s EEG, one quasi-brain-death patient’s EEG, and one special patient’s EEG which was from coma to quasi-brain-death state were processed to present individual characteristics. Results show intuitively that patients in coma state have higher dynamic EEG energy and lower complexity distribution than patients in quasi-brain-death state.