Classification of neurological and mental health disorders based on multimodal approaches: A comprehensive review

Abstract

Disorders of the nervous system and mental health are among the most prevalent, complex, and devastating health challenges globally, with a significant impact on quality of life. Recently, advances in deep learning-based multimodal methodologies have transformed the classification and detection of these disorders through the utilization of diverse data types including neuroimaging, bio-signals, and clinical evaluations. These multimodal techniques provide a more holistic understanding of complex conditions, addressing the limitations of traditional unimodal methods, which often fail to capture the multifaceted nature of these disorders. Despite the growing body of research, a comprehensive review focusing on the application of deep learning-based multimodal approaches to both neurological and mental health disorders remains lacking. This review fills this gap by offering an in-depth analysis of recent advancements in machine learning and deep learning-based multimodal classification for ten major disorders: five neurological and five mental health-related. It examined key modalities, explored fusion strategies, and provided insights into the strengths and weaknesses of existing multimodal approaches. Additionally, this review highlighted the challenges associated with multimodal data integration, such as data imbalance, model interpretability, and the need for large-scale, high-quality datasets. Furthermore, the review discussed emerging trends and future directions, emphasizing the potential of advanced fusion and computational techniques to enhance the clinical applicability of these models. By synthesizing the current state of research, this review aims to guide future studies and contribute to the development of more accurate, reliable, and accessible diagnostic tools for neurological and mental health disorders.