A Study of Informative EEG Channel and Brain Region for Typing Activity

Abstract

Electroencephalography (EEG) is an electrophysiological monitoring method to record electrical activity of the brain. Brain consists of four lobes which is frontal, parietal, temporal and occipital lobe. Each lobe has their own respective function and it release different rhythmic wave when carry out different actions. However, when performing a specific activity, not all the EEG channels tend to be informative to the particular activity. The need to optimize the number of channels is crucial to reduce computational complexity. The aim of this paper is to determine the informative EEG channel/s and brain region for typing activity. 20 healthy with right-handed subjects from Universiti Malaysia Perlis (UniMAP) were enrolled in this study. Typing task was performed for 3 trials and 5 minutes per trial. In EEG signal processing, Notch filter and Butterworth bandpass filter were used to remove powerline artefact and to filter the signal into alpha (8-13Hz) and beta waves (13-30Hz). Welch method was applied to extract features from typing task. The obtained results were then undergoing the statistical analysis before load into the K-Nearest Neighbour (KNN) and Linear Discriminant Analysis (LDA) classifier. Based on this study, it is found that channel P3 in parietal region and channel T6 in temporal region give highest accuracy which is 99.44% for typing task activity.