Miniformer: A Minimalist Transformer for Brain Functional Networks Analysis
- 21 May 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Biomedical and Health Informatics
- Vol. 29 (11) , 8460-8473
Abstract
Learning to estimate and classify brain functional networks (BFNs) has become an increasingly important way of predicting neurological or mental disorders at their early stages. The traditional methods conduct BFN estimation and classification in two separate steps, thus preventing the interaction and joint optimization. In contrast, Transformer provides a natural architecture to learn BFNs with downstream tasks in an end-to-end manner. Despite their great potential, Transformer-based methods involve a large number of parameters that need to be learnt from big data and often lead to poor model interpretability. Considering the challenge in acquiring data and the high demand for model interpretability in medical scenarios, in this paper, we propose a minimalist Transformer architecture, referred to as Miniformer, by simplifying the projection matrices in the self-attention module into a single diagonal matrix, which greatly reduces the number of parameters, alleviates the risk of overfitting, and improves the interpretability. Additionally, the clear physical meaning of parameters in Miniformer makes the integration of domain knowledge or prior easier and more natural. Therefore, we further develop two variants of Miniformer by incorporating sparsity for removing potentially noisy time points from fMRI signals, and smoothness for capturing the temporal correlations in fMRI signals, respectively. To evaluate the effectiveness of the proposed methods, we perform brain disease diagnosis experiments on three public datasets. The results show that Miniformer and its variants tend to achieve higher classification performance than comparison methods with good interpretability.Keywords
Funding Information
- National Natural Science Foundation of China (62476155, 62176112)
- Natural Science Foundation of Shandong Province (ZR2024MF063)
This publication has 58 references indexed in Scilit:
- Modularity-Constrained Dynamic Representation Learning for Interpretable Brain Disorder Analysis with Functional MRI2023
- Unsupervised contrastive graph learning for resting‐state functional MRI analysis and brain disorder detectionHuman Brain Mapping, 2023
- Functional brain network based multi-domain feature fusion of hearing-Impaired EEG emotion identificationBiomedical Signal Processing and Control, 2023
- Abnormal core functional connectivity on the pathology of MDD and antidepressant treatment: A systematic reviewJournal of Affective Disorders, 2022
- A toolbox for brain network construction and classification (BrainNetClass)Human Brain Mapping, 2020
- Specific Functional Connectivity Patterns of Middle Temporal Gyrus Subregions in Children and Adults with Autism Spectrum DisorderAutism Research, 2019
- Inter-Network High-Order Functional Connectivity (IN-HOFC) and its Alteration in Patients with Mild Cognitive ImpairmentNeuroinformatics, 2019
- Extraction of dynamic functional connectivity from brain grey matter and white matter for MCI classificationHuman Brain Mapping, 2017
- Automated Anatomical Labeling of Activations in SPM Using a Macroscopic Anatomical Parcellation of the MNI MRI Single-Subject BrainNeuroImage, 2002
- Functional connectivity in the motor cortex of resting human brain using echo‐planar mriMagnetic Resonance in Medicine, 1995