Stable individualized brain computing model informed by spatiotemporal co-activity patterns.

Area of Science:

• Neuroscience
• Computational Biology
• Brain Network Modeling

Background:

• Accurate brain simulation is crucial for understanding cognition, behavior, and developing personalized therapies for brain diseases.
• Traditional neurodynamics models, relying on structural connectivity, struggle to fully capture brain information for high-fidelity individual-level simulations.
• Existing models face challenges in achieving accurate whole-brain activity simulations at the individual level.

Purpose of the Study:

• To introduce a novel data-driven framework, the Stable Individualized Brain Computing Model (SI-BCM), for simulating whole-brain activity.
• To infer spatiotemporal co-activity patterns from fMRI data to capture intrinsic functional collaboration patterns.
• To enhance the accuracy and reliability of individual-level brain simulations.

Main Methods:

• Developed the Stable Individualized Brain Computing Model (SI-BCM), a data-driven reverse engineering framework.
• Integrated spatiotemporal dimensional information to extract stable and shared connectivity patterns representing intrinsic functional collaboration.
• Incorporated a novel cost function based on the Phase-Space Association (PSA) matrix to improve dynamics capture.
• Utilized functional magnetic resonance imaging (fMRI) data for inferring whole-brain activity patterns.

Main Results:

• Achieved a high correlation coefficient of 0.87 between simulated and empirical functional connectivity (FC).
• Demonstrated enhanced simulation accuracy, robustness, and reliability at the individual level compared to existing models.
• Showcased the model's sensitivity to changes in cognitive function, offering insights into neural mechanisms.
• Successfully applied SI-BCM to model Alzheimer's disease (AD) patients, supporting the hypothesis of excessive neuronal excitation in AD pathogenesis.

Conclusions:

• The SI-BCM establishes a new paradigm for brain network modeling by prioritizing stable dynamics inference from activity data.
• This framework provides a powerful tool for understanding complex brain function and pathophysiology.
• The model's success in simulating individual brain activity and modeling AD offers significant potential for clinical applications and research.