Deep learning and whole-brain networks for biomarker discovery: modeling the dynamics of brain fluctuations in resting-state and cognitive tasks

2024-12-26

AI Summary

This study investigates whether whole-brain dynamical bifurcation parameters can serve as novel, interpretable biomarkers to distinguish resting-state from task-evoked brain functional states and assess their diagnostic potential in neuropsychiatric disorders. Method: Leveraging a supercritical Hopf model to generate synthetic BOLD signals, we invert key bifurcation parameters from multimodal fMRI data of the Human Connectome Project (HCP) using a hybrid CNN–RNN deep learning framework. Contribution/Results: We provide the first systematic validation that the bifurcation parameter is significantly elevated during task performance versus rest (all tasks vs. rest: *p* < 0.0001) and scales monotonically with cognitive load. The parameter exhibits robust discriminative performance and strong cross-subject generalizability. Critically, it constitutes a mechanistically grounded, quantifiable functional brain state metric—bridging computational neurodynamics and clinical neuroimaging biomarker discovery through a novel, theory-informed paradigm.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Machine Learning: Bio-inspired Learning
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Background: Brain network models offer insights into brain dynamics, but the utility of model-derived bifurcation parameters as biomarkers remains underexplored. Objective: This study evaluates bifurcation parameters from a whole-brain network model as biomarkers for distinguishing brain states associated with resting-state and task-based cognitive conditions. Methods: Synthetic BOLD signals were generated using a supercritical Hopf brain network model to train deep learning models for bifurcation parameter prediction. Inference was performed on Human Connectome Project data, including both resting-state and task-based conditions. Statistical analyses assessed the separability of brain states based on bifurcation parameter distributions. Results: Bifurcation parameter distributions differed significantly across task and resting-state conditions ($p<0.0001$ for all but one comparison). Task-based brain states exhibited higher bifurcation values compared to rest. Conclusion: Bifurcation parameters effectively differentiate cognitive and resting states, warranting further investigation as biomarkers for brain state characterization and neurological disorder assessment.

Problem

• Brain Network Models
• Resting vs Thinking State Discrimination
• Brain Disease Diagnosis

Innovation

• Deep Learning
• Whole Brain Network Analysis
• Brain State Distinction