Analysis of EEG data room in sudden infant death risk patients

A novel method for EEG feature analysis applying neural networks and evolut ionary algorithms is proposed. Polysomnographic data of neonate infants are analyzed in order to detect characteristic EEG signatures that correlate w ith a certain apnea risk group. The method described works without preselected categories, showing the poss ibility of a purely EEG-based detection of sudden infant death risk patient s. Our method employed the registration of relative distance functions and trend analysis of variable EEG features. It is thus shown that an adaptive hyperplane within the multidimensional EEG data room successfully converges towards a unique feature vector that represents the relative EEG character istics within the patient pool.