UNVEILING AUTISM EARLY: A HOLISTIC REVIEW OF DETECTION STRATEGIES

Abstract

Autism spectrum disorder (ASD) is a neurodevelopmental disorder caused by environmental and genetic factors. ASD significantly impacts on social communication, interaction, and behavior. Therefore, early and accurate diagnosis is essential for timely intervention, which can improve outcomes by enabling tailored therapeutic strategies to deal with such disease. Traditional diagnostic approaches rely on behavioral assessments, which can be subjective, time-consuming, and resource-intensive. To address these limitations, researchers have explored automated diagnostic methods using artificial intelligence techniques, particularly machine learning and deep learning. This review presents a comprehensive analysis of ASD detection strategies that utilize various data modalities, including facial analysis, retinal imaging, electroencephalography (EEG), and eye-tracking, that are captured from ASD patients. Each modality offers unique insights into ASD characteristics, with AI-based models demonstrating promising results in distinguishing autistic from non-autistic individuals Given the improvement in diagnostic methods and algorithms for ASD, we examined research articles from the last decade (2015–2025) that highlight recent contributions in identifying ASD in children. In addition, this review outlines the key features of each approach, summarizes their results, and discusses the challenges they present.