Evaluation Between Visual And Auditory EEG Responses On Upper Limb Movement in The Temporal and Spectral Domains

Abstract

Visual and auditory stimulus modalities have been used widely in designing EEG experimental protocols. The influence of the visual and auditory stimulus modalities on the efficiency of signal classifiers is yet to be thoroughly assessed and documented in the context of motor movement prediction using EEG signals. Therefore, finding a suitable choice of stimulus modality in designing EEG experiments could improve the expected outcome of the research. In this paper, we aim to investigate the effect of two stimulus modalities, vision and sound, based on EEG signal features of different motor execution for finding the better stimulus modalities. We collected EEG data on 11 subjects qualified for inspection, giving over 2,750 motor execution samples per experiment. The subjects participated in 2 experimental protocols, which were built for explicitly performing the flexion and extension of the left and right elbow simultaneously. Two features in the temporal domain were computed, and a fast Fourier transform algorithm was applied to produce time-frequency domain scalograms for comparison between left and right elbow movements from visual and audio stimulated experiments. Our research showed high feasibility of classifying motor movements based on 2 time-domain features: the entropy and upper quartile, and the Fourier time-frequency domain scalogram in the power region between 4–16 Hz. In the time-frequency domain, the density of high power region spectra with visual stimuli is greater than that of auditory stimuli.