Self-Supervised Learning to Unveil Brain Dysfunctional Signatures in Brain Disorders: Methods and Applications

Abstract

Importance: Precisely decoding brain dysfunction from high-dimensional functional recordings is crucial for advancing our understanding of brain dysfunction in brain disorders. Self-supervised learning (SSL) models offer a transformative approach for mapping dependencies in functional neuroimaging data. Leveraging the intrinsic organization of brain signals for comprehensive feature extraction, these models enable the analysis of critical neurofunctional features within a clinically relevant framework, overcoming challenges related to data heterogeneity and the scarcity of labeled data. Highlight: This paper provides a comprehensive overview of SSL techniques applied to functional neuroimaging data, such as functional magnetic resonance imaging and electroencephalography, with a specific focus on their applications in various neuropsychiatric disorders. We discuss 3 main categories of SSL methods: contrastive learning, generative learning, and generative-contrastive methods, outlining their basic principles and representative methods. Critically, we highlight the potential of SSL in addressing data scarcity, multimodal integration, and dynamic network modeling for disease detection and prediction. We showcase successful applications of these techniques in understanding and classifying conditions such as Alzheimer's disease, Parkinson's disease, and epilepsy, demonstrating their potential in downstream neuropsychological applications. Conclusion: SSL models provide a scalable and effective methodology for individual detection and prediction in brain disorders. Despite current limitations in interpretability and data heterogeneity, the potential of SSL for future clinical applications, particularly in the areas of transdiagnostic psychosis subtyping and decoding task-based brain functional recordings, is substantial.

Fig. 1.

Overview of the typical SSL pipeline for neuroimaging data analysis. The top represents the brain network pipeline, where raw neurological data are systematically processed to extract meaningful representations. The bottom highlights the core self-supervised model, comprising an encoder–decoder architecture. These refined representations are then utilized for downstream tasks, such as disease categorization, detection, and prediction. The model’s bidirectional learning flow ensures robustness and adaptability across diverse neuroimaging datasets.

Fig. 2.

The primary learning strategies within SSL models in neuroimage-based medical applications. In contrastive learning, the graph-based approach generates augmented views of brain graphs to maximize view similarity through encoders and decoders, while the spatiotemporal-based approach focuses on leveraging temporal neural signals for similar contrastive objectives. Generative learning includes a mask-based method, which reconstructs randomly masked brain regions to minimize reconstruction loss, and a VAE-based method, where neural imaging data are encoded and reconstructed to learn global patterns. Last, generative-contrastive learning combines generative modeling, such as GANs, with contrastive learning to capture intrinsic brain representations.