Investigation and analysis of brain signals for correlation of eye blink rate and event related potential for cognitive evaluation

Abstract

The quantitative assessment of the cognitive capacity of an individual is a critical factor in diagnosing diseases that cause cognitive impairment and for analysing the rate of progression. The conventional cognitive tests use questionnaire - based evaluation which is followed by the analysis of ElectroEncephaloGram (EEG) to provide a quantitative measure of the decline of cognition in different levels. The present research work has focused on bringing out a method for evaluating the cognitive capacity of an individual using Eye Blink Rate (EBR) obtained from ElectroOculoGram (EOG) and, P300 which is an event related component of the EEG using Brain Computer Interface (BCI) prediction system. The first phase of work was carried out to find out the selection of electrode position for the predominant occurrence of eye blink based on the International standard 10-20 electrode system. An automated algorithm, with the EOG signal as input for identification of eye blink based on threshold parameter, was developed. Using Support Vector Machine (SVM) classifier with Training Feature Matrix (TFM) in MATLAB was used to classify cognitive impairment. To enhance this process, along with the eye blink from EOG, another parameter P300 from EEG was included for cognitive prediction system. P300 contain high information and also requires minimum user training. Here, a new method is presented for representing the EEG/EOG signal according to peak amplitude present in it. newline