Robust thalamic nuclei segmentation method based on local diffusion magnetic resonance properties.

Area of Science:

• Neuroimaging
• Neuroanatomy
• Diffusion MRI

Background:

• The thalamus, a key relay in brain connections, has a complex structure of nuclei crucial for neurological functions.
• Accurate segmentation of thalamic nuclei is vital for understanding brain connectivity and function.

Purpose of the Study:

• To develop and validate a novel framework for segmenting thalamic nuclei using diffusion magnetic resonance imaging (dMRI).
• To improve the differentiation of intra-thalamic microstructure by leveraging orientation distribution functions (ODFs) and spherical harmonic (SH) representations.

Main Methods:

• Utilized dMRI data at 3 Tesla to explore ODFs, enhanced by SH representation for detailed angular characterization.
• Employed k-means clustering, initialized in a data-driven manner, to segment thalamic nuclei.
• Validated the method on 35 healthy volunteers, comparing results to Morel's histological atlas and assessing reproducibility across scanners and sessions.

Main Results:

• Achieved robust, reproducible, and accurate segmentation of the thalamus into seven nuclei groups.
• Six segmented groups corresponded closely to known anatomical labels (anterior, ventral anterior, medio-dorsal, ventral latero-ventral, ventral latero-dorsal, pulvinar).
• The seventh cluster comprised centro-lateral and latero-posterior nuclei, showing overlap with major white matter tracts passing through the thalamus.

Conclusions:

• The proposed SH-based ODF framework provides superior segmentation accuracy compared to previous state-of-the-art methods.
• This anatomically reliable segmentation of thalamic nuclei holds significant potential for clinical applications in neurology.
• The method demonstrates robustness across different MRI sequences and scanners, ensuring broad applicability.