Automatic Judgment of Open/Closed Eye States for Accurate Interpretation of Awake Background EEG

Abstract

Automatic judgment of open/closed eye states from the awake background electroencephalogram (EEG) recording has been always in a high demand for neurological signal analysis and plays an important role for posterior dominant rhythm (PDR) analysis. PDR appears predominantly in occipital lobes and contains significant information for interpreting fundamental brain dysfunctions. The aim of this research is to develop a system that can properly differentiate between open and closed eye states, so that some specific segments can be chosen for PDR analysis that appears just after eye open/close. In this proposed method, a computer assisted automatic system for eye-opening/closing detection from awake background EEG has been developed. EEG data that was visually inspected by a qualified electroencephalographer (EEGer) was taken into account for separating open and closed eye states by creating parameters for each states. Later using those parameters and conditions, new equations were developed and implemented for accurate detection of open/closed eye. Based on the automatic detection result, some specific segments that appears just after eye open/close will be selected for PDR analysis. Organization, frequency, amplitude and their asymmetry these characteristics will be taken into account for PDR analysis.