Bio-signals based automatic sleep-stage identification system

摘要

The present invention relates to a bio-signal-based automatic sleep stage classification system, the bio-signals of different conditions is to estimate the surface of the water phase using. To this end, the present invention is EEG, EOG, EMG, PPG, SKT, using a GSR sensor different channels to different bio bio-signal detector to acquire signals using various biological signals obtained through the respective channel in each of the biological signal detection section a feature point for each of the signal processing and feature extraction unit, a signal processing and feature extraction unit for extracting feature points of each physiological signal indicating the characteristics of the bio-signal from bio-signal subjected to signal processing by the feature extraction method for the bio-signal subset generation section for generating, based on the bio-signal combining, and the sleep stage classification (REM, LS, DS) through the combination of bio-signals from the combined subset generation unit, and the performance evaluation of analyzing the performance evaluation of the combination according to the steps by providing a bio-signal-based automatic sleep stage classification system including a, can correct the sleep phase at the best bio-signal combining considering the accuracy and ease of measurement during sleep phase analysis using the bio-signal of the different conditions and easily predictable allows. ;