Direct diffusion-based parcellation of the human thalamus

Area of Science:

• Neuroimaging
• Human Anatomy
• Diffusion Tensor Imaging

Background:

• The human thalamus is a complex structure with diverse functions.
• Understanding its stable anatomical subdivisions and connectivity is crucial for neuroscience.
• Previous parcellation methods have limitations in unbiased segmentation.

Purpose of the Study:

• To identify and characterize stable, reproducible thalamic substructures using an unbiased approach.
• To determine the localization, size, and individual variations of these thalamic clusters.
• To investigate the subcortical and cortical connectivity patterns of identified thalamic clusters.

Main Methods:

• Employed an unbiased diffusion tensor parcellation approach for thalamic segmentation.
• Analyzed data from 63 healthy human subjects (32 males, 31 females).
• Selected five stable thalamic clusters common to 90% of subjects for detailed analysis.

Main Results:

• Identified 21 thalamic clusters, with laterality and gender differences accounting for ±6% and ±4% of thalamic volume, respectively.
• Characterized five stable clusters: anteroventral nucleus (AN) group, mediodorsal (MD) nucleus, medial pulvinar (PuM), and two lateral nuclei groups.
• Revealed distinct connectivity: AN to limbic structures, MD to prefrontal cortex/medial temporal lobe, PuM to visual/medial temporal areas, and lateral clusters to sensory-motor cortex.

Conclusions:

• Diffusion tensor parcellation successfully identified stable thalamic substructures with consistent anatomical features.
• The identified clusters (AN, MD, PuM, lateral) exhibit specific, reproducible connectivity patterns.
• These findings provide a detailed map of thalamic organization and its role in connecting diverse brain regions.