A Review on Attention-Driven Multimodal Deep Learning Techniques for Early Detection of Autism Spectrum Disorder Using Brain Imaging Data

Machine learning

Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition that affects communication, behavior, and social interaction. Early diagnosis is essential for effective intervention, but traditional clinical methods are often subjective and time-consuming. This paper presents a comprehensive review of attention-driven multimodal deep learning techniques for the early detection of ASD using neuroimaging data such as MRI and fMRI. Deep learning models, particularly Convolutional Neural Networks (CNNs), are effective in extracting meaningful features from brain images, while attention mechanisms enhance model performance by focusing on relevant brain regions. Multimodal learning, which integrates structural and functional imaging data, further improves diagnostic accuracy and reliability. The study also discusses commonly used datasets, performance metrics, challenges such as data imbalance and high computational cost, and emerging trends including transformer models and explainable AI. Overall, this review highlights the potential of advanced AI techniques in improving ASD diagnosis and supporting clinical decision-making.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.