Adaptive Whole-Brain Dynamics Predictive Method: Relevancy to Mental Disorders

Area of Science:

• Computational neuroscience
• Neuroimaging analysis
• Clinical neurology

Background:

• Traditional brain connectivity models lack precision in capturing dynamic brain states.
• Parameter fitting in brain models is often imprecise, limiting clinical applications.
• Heterogeneity parameters are crucial for understanding dynamic brain characteristics.

Purpose of the Study:

• To refine parameter fitting methods for the Hopf whole-brain model.
• To improve the accuracy and stability of dynamic brain characteristic quantification.
• To identify neuropathological differences in major depressive disorder (MDD) and autism spectrum disorder (ASD).

Main Methods:

• Validated parameter fitting using simulated and synthetic networks.
• Introduced individual-specific initialization and optimized gradient descent.
• Developed an approximate loss function and gradient adjustment mechanism.
• Applied the refined model to MDD and ASD patient datasets.

Main Results:

• Enhanced parameter fitting accuracy and stability.
• Identified distinct brain region differences between patients with MDD/ASD and healthy controls.
• Successfully explained anomalies associated with these neuropsychiatric disorders.

Conclusions:

• The refined Hopf whole-brain model offers precise neuropathological identification.
• This approach has substantial potential for novel treatments in neuropsychiatric research.
• Validated methods are crucial for the clinical application of brain modeling.