Rajagopalan Shyam Sundar, Ghosh Sanjay
Institute of Bioinformatics and Applied Biotechnology, Bengaluru, India.
Methods Mol Biol. 2025;2952:233-242. doi: 10.1007/978-1-0716-4690-8_13.
Autism Spectrum Disorder (ASD or Autism) is a neurodevelopmental disorder that is characterized by challenges in social communication skills and the presence of restricted and repetitive behaviors during early childhood. ASD poses a significant public health challenge with increasing prevalence rates worldwide. Early diagnosis and intervention are critical for improving outcomes in children with ASD. However, current diagnostic methods often involve subjective assessments and are time-consuming. Currently, there are no known biomarkers for ASD, and the diagnosis is based on phenotypic manifestations observed by trained clinicians over time. Additionally, the heterogeneity of Autism and associated co-occurring conditions pose further challenges for screening and early detection. Recent advances in Artificial Intelligence (AI) and Machine Learning (ML) are transforming ASD screening and diagnosis. These computational technologies are capable of analyzing complex datasets and multiple modalities, including multi-omics, brain images, behavior assessments, medical and background information, and registry data to identify patterns that may not be evident to clinicians or parents. This article reviews recent developments in the application of AI/ML for ASD screening and early diagnosis. It also covers the use of AI/ML in understanding the biological underpinnings of ASD.
自闭症谱系障碍（ASD 或自闭症）是一种神经发育障碍，其特征是在幼儿期社交沟通技能方面存在挑战，以及出现受限和重复行为。随着全球患病率的上升，ASD 对公共卫生构成了重大挑战。早期诊断和干预对于改善自闭症谱系障碍儿童的预后至关重要。然而，目前的诊断方法通常涉及主观评估且耗时。目前，尚无已知的自闭症谱系障碍生物标志物，诊断基于训练有素的临床医生长期观察到的表型表现。此外，自闭症的异质性以及相关的共病情况给筛查和早期检测带来了进一步的挑战。人工智能（AI）和机器学习（ML）的最新进展正在改变自闭症谱系障碍的筛查和诊断。这些计算技术能够分析复杂的数据集和多种模式，包括多组学、脑图像、行为评估、医疗和背景信息以及登记数据，以识别临床医生或家长可能不明显的模式。本文综述了人工智能/机器学习在自闭症谱系障碍筛查和早期诊断应用中的最新进展。它还涵盖了人工智能/机器学习在理解自闭症谱系障碍生物学基础方面的应用。
