Foundation models for discovering robust biomarkers of neurological disorders from dynamic functional connectivity

2026-04-23

AI Summary

Although existing brain foundation models can predict neurological disorders, the biomarkers they identify often lack robust validation. To address this limitation, this work proposes the RE-CONFIRM framework, which systematically evaluates the stability of biomarkers derived from dynamic functional connectivity for the first time. The study introduces Hub-LoRA, a novel fine-tuning strategy that enhances the neurobiological plausibility of identified biomarkers while preserving model efficiency. Evaluated across five fMRI datasets encompassing autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), and Alzheimer’s disease, Hub-LoRA not only outperforms specialized deep learning models in predictive performance but also yields robust biomarkers that align closely with established neural mechanisms, thereby significantly improving interpretability and clinical translatability.

Technology Category

• Machine Learning: Neuro-Symbolic Learning
• Cognitive Modeling & Cognitive Systems: Other Foundations of Cognitive Modeling & Systems
• Natural Language Processing: Safety and Robustness

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Several brain foundation models (FM) have recently been proposed to predict brain disorders by modelling dynamic functional connectivity (FC). While they demonstrate remarkable model performance and zero- or few-shot generalization, the salient features identified as potential biomarkers are yet to be thoroughly evaluated. We propose RE-CONFIRM, a framework for evaluating the robustness of potential biomarker candidates elucidated by deep learning (DL) models including FMs. From experiments on five large datasets of Autism Spectrum Disorder (ASD), Attention-deficit Hyperactivity Disorder (ADHD), and Alzheimer's Disease (AD), we found that although commonly used performance metrics provide an intuitive assessment of model predictions, they are insufficient for evaluating the robustness of biomarkers identified by these models. RE-CONFIRM metrics revealed that simply finetuning FMs leads to models that fail to capture regional hubs effectively, even in disorders where hubs are known to be implicated, such as ASD and ADHD. In view of this, we propose Hub-LoRA (Low-Rank Adaptation) as a fine-tuning technique that enables FMs to not only outperform customised DL models but also produce neurobiologically faithful biomarkers supported by meta-analyses. RE-CONFIRM is generalizable and can be easily applied to ascertain the robustness of DL models trained on functional MRI datasets. Code is available at: https://github.com/SCSE-Biomedical-Computing-Group/RE-CONFIRM.

Problem

• biomarkers
• dynamic functional connectivity
• foundation models
• robustness evaluation
• neurological disorders

Innovation

• foundation models
• dynamic functional connectivity
• biomarker robustness
• Hub-LoRA
• RE-CONFIRM