Revisiting the Role of Structural Connectivity-Based Parcellation in Thalamic Nuclei Segmentation: comparison with...

Area of Science:

• Neuroimaging
• Neuroscience
• Anatomy

Background:

• Accurate thalamic nuclei segmentation is vital for neuroscience research and clinical interventions like deep brain stimulation.
• Connectivity-based parcellation (CBP) is widely used but its anatomical validity is uncertain compared to newer methods.

Purpose of the Study:

• To compare the anatomical accuracy of CBP with other state-of-the-art segmentation methods.
• To evaluate the performance of different diffusion and structural MRI techniques for thalamic nuclei segmentation.

Main Methods:

• Analysis of high-resolution diffusion MRI and T1-weighted data from the Human Connectome Project (n=67).
• Connectivity-based parcellation using probabilistic tractography with cortical targets.
• Comparison with orientation distribution function (ODF) clustering, track density imaging (TDI), and structural MRI segmentation.
• Group-level analysis in MNI space and Dice overlap calculation against the Morel atlas.

Main Results:

• CBP showed limited anatomical precision, with increased variability and noise at higher target counts.
• ODF clustering and TDI revealed subdivisions consistent with cytoarchitectonic patterns, especially in pulvinar and mediodorsal nuclei.
• Structural MRI segmentation achieved the highest Dice coefficients, closely matching Morel atlas boundaries; CBP consistently underperformed.

Conclusions:

• CBP is limited in delineating thalamic nuclei with histological accuracy.
• Structural and diffusion microstructural (ODF, TDI) approaches offer superior nuclear localization.
• Hybrid workflows integrating structural and diffusion information are needed for reliable thalamic segmentation in research and clinical applications.