Advancements in EEG-Based Machine Learning Techniques for Early Autism Spectrum Disorder Diagnosis: A Review

Abstract

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition characterized by impairments in social interaction, communication, and repetitive behaviors. Electroencephalography (EEG) has gained prominence as a reliable tool for ASD diagnosis, capturing critical behavioral patterns. Researchers have applied various Machine Learning (ML) techniques to enhance ASD detection, achieving notable accuracy. Studies using feature selection with ML classifiers have reported up to 100% accuracy in children as young as 6–12 months. Other approaches integrating EEG with behavioral features such as eye gaze, facial gestures, and body movements have attained classification accuracies as high as 87.5%. Additionally, resting-state EEG studies have explored microstate differences between ASD and neurotypical individuals. Several ML models, including Support Vector Machines (SVM), Convolutional Neural Networks (CNNs), and ensemble methods like Random Forest and Naïve Bayes, have demonstrated classification precision between 90% and 99%. However, challenges such as data heterogeneity and limited sample sizes hinder clinical implementation. This review highlights the most notable EEG-based ML studies for ASD diagnosis and emphasizes the need for further research to refine these techniques for broader clinical adoption.