EEG Classification Analysis for Diagnosing Autism Spectrum Disorder based on Emotions

Abstract

This research sets out to propose another method for the medical teams to diagnose Autism Spectrum Disorder (ASD) in children based on the analysis on Electroencephalography (EEG). Its main intention is to provide an effective and more time-saving method in diagnosing autism in suspected autistic children and to choose the best technique for classifying the EEG data to distinguish the Autistic traits from the normal ones. This research paper comprises of EEG data analysis on the brainwave activities of normal individuals and autism subjects to learn and compare the difference between their brain activity’s patterns. In addition, classification and analysis were also done to distinguish the different emotion projection of autistic and normal subjects and how their characteristics differ from each other. The signal processing techniques were performed on EEG data obtained from chosen subjects and two of the most well-known machine learning techniques which are Artificial Neural Network (ANN) and Support Vector Machine (SVM) were utilized in classifying the different classes of brainwave activities and signals. The efficiency of the two classifiers was then compared. For ANN, an experiment to determine the optimum value of hidden layers was also executed. The results obtained from this research provided classification accuracy that can be obtained from normal and autistic data classification as well as the ability to diagnose a new data using the trained ANN. Positive findings were obtained from this EEG analysis especially in classifying normal and Autistic patterns and also in classifying the different emotions. This outcome can very much help in the process of diagnosing ASD, where the whole process can be done in a more time-efficient manner and more accurate diagnosis can be made.