Working memory impairments imitate age-related behaviors in children using visual stimulation based on event-related potentials

Abstract

The aim of this study is to examine the working memory impairments imitate age-related between 7 to 12 years old using Event-Related Potentials (ERP) signal. 97 normal children were selected to a visual stimuli assessment (Phase 1 and Phase 2) while their working memory response was recorded using Electroencephalograph (EEG) machine. Raw EEG signal were segmented and averaged into the ERP signal according to the event stimulus occur. Discrete Wavelet Transform technique is preferred to decompose the ERP signal into different frequency band. ERP signal at alpha frequency is used because of alpha is the most prominent component of brain waves activity. The necessary features were extracted as an input for the Logistic Regression (LR) and Support Vector Machine (SVM) classifier. Consequence indicated that the accuracy and mean performance results were significant in predicting either a child had working memory impairment or not. 7 years old have lower accuracy compared to other groups with 60% for LR and 86% for SVM. In conclusion, the study proposed that age-related changes and increasing level of visual stimuli affect working memory impaired. Thus, this study has provided empirical evidence in support for the assumption that younger children have working memory impaired through visual stimuli assessment.