Reproducibility of thalamic segmentation based on probabilistic tractography.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Neurosurgery

Background:

• Accurate identification of thalamic nuclei is critical for effective Deep Brain Stimulation (DBS) electrode placement.
• Understanding thalamic function in health and disease necessitates precise segmentation of its nuclei.

Purpose of the Study:

• To implement and assess a probabilistic tractography method for thalamic nuclei segmentation.
• To evaluate the reproducibility of this segmentation method both within and between subjects.
• To investigate the impact of using different numbers of cortical target regions (6 vs. 31) on segmentation reproducibility.

Main Methods:

• Probabilistic tractography was employed to segment thalamic nuclei based on their connections to cortical regions.
• Quantitative assessment of segmentation overlap was used to measure within- and between-subject reproducibility.
• The study compared segmentation results using 6 versus 31 cortical target regions.

Main Results:

• Very high reproducibility was achieved when processing a single dataset multiple times.
• Thalamic segmentation demonstrated high reproducibility across multiple datasets from the same subject using 6 target regions.
• Increasing target regions to 31 reduced within-subject reproducibility, especially in medial and posterior thalamus.
• Significant overlap in segmentation results was observed between different subjects, particularly for regions connected to frontal, parietal, temporal, and pre-central cortices.

Conclusions:

• The implemented probabilistic tractography method provides a highly reproducible approach for thalamic nuclei segmentation.
• Using 6 cortical target regions offers a good balance between reproducibility and anatomical detail for thalamic segmentation.
• The findings support the utility of this method for improving DBS targeting and advancing thalamic research.