Thalamus Parcellation Based on Evidence Accumulation Clustering Using dMRI Tractography

Area of Science:

• Neuroscience
• Computational Neuroscience
• Medical Imaging

Background:

• The thalamus is a critical brain structure for sensory, motor, and cognitive functions.
• Accurate thalamus parcellation is essential for understanding brain function and neurological disorders.
• Current diffusion MRI (dMRI) tractography methods for thalamus parcellation lack reliability and stability due to single-pipeline reliance.

Purpose of the Study:

• To develop a novel and robust framework for thalamus parcellation.
• To overcome the limitations of single-pipeline dMRI tractography methods.
• To improve the accuracy and stability of thalamic subregion localization.

Main Methods:

• Proposed a novel thalamus parcellation framework utilizing Evidence Accumulation Clustering (EAC).
• Integrated results from multiple dMRI processing pipelines to enhance robustness.
• Applied the EAC framework to the Human Connectome Project (HCP) test-retest dataset.

Main Results:

• The proposed EAC-based method demonstrated superior stability and accuracy in localizing thalamic subregions.
• Outperformed traditional single-pipeline dMRI tractography approaches.
• Mitigated systematic errors and incorporated diverse structural connectivity information.

Conclusions:

• The EAC framework provides a more reliable and accurate method for thalamus parcellation.
• This advancement has implications for cognitive neuroscience research and clinical applications in neurological disorders.
• The integrated multi-pipeline approach enhances the understanding of complex thalamic subregion structures.