Brain Effective Connectivity Modeling for Alzheimer's Disease by Sparse Gaussian Bayesian Network

Area of Science:

• Neuroscience
• Computational Biology
• Medical Imaging

Background:

• Alzheimer's disease (AD) is linked to altered brain connectivity networks.
• Effective connectivity, the directional relationship between brain regions, is crucial for brain function.
• Limited research exists on modeling AD's effective connectivity and its distinction from normal controls (NC).

Purpose of the Study:

• To investigate sparse Bayesian Networks (BNs) for effective connectivity modeling in Alzheimer's disease.
• To propose a novel BN structure learning formulation with L1-norm and directed acyclic graph (DAG) constraints.
• To compare the proposed method's accuracy and scalability against existing algorithms.

Main Methods:

• Developed a novel sparse Bayesian Network structure learning algorithm incorporating L1-norm and DAG constraints.
• Conducted theoretical analysis and experiments on benchmark networks to validate the method's performance.
• Applied the method to FDG-PET images from 42 AD and 67 NC subjects to model effective connectivity.

Main Results:

• The proposed sparse BN method demonstrated superior learning accuracy and scalability compared to ten competing algorithms.
• Identified distinct effective connectivity models for Alzheimer's disease (AD) and normal controls (NC).
• Observed significant differences in global, intra-lobe, inter-lobe, and inter-hemispheric effective connectivity between AD and NC groups.

Conclusions:

• The novel sparse Bayesian Network approach effectively models and differentiates brain effective connectivity in Alzheimer's disease.
• Findings align with known AD pathology and clinical progression, offering insights into disease mechanisms.
• This study contributes to AD knowledge discovery by characterizing network alterations in the disease.