Subcortical gray matter segmentation and voxel-based analysis using transverse relaxation and quantitative...

Area of Science:

• Neuroimaging
• Medical Physics
• Radiology

Background:

• Multiple Sclerosis (MS) affects the central nervous system, often involving deep gray matter structures.
• Accurate segmentation and analysis of deep gray matter are crucial for understanding MS pathophysiology.
• Quantitative susceptibility mapping (QSM) and transverse relaxation rate (R2 *) offer novel insights into tissue properties.

Purpose of the Study:

• To investigate subcortical gray matter segmentation using R2 * and QSM.
• To apply these quantitative MRI techniques to voxel-based analysis in multiple sclerosis (MS).

Main Methods:

• Voxel-based analysis of R2 * and QSM in deep gray matter nuclei (caudate, putamen, globus pallidus, thalamus).
• Comparison with standard whole-structure analysis in 37 MS subjects and 37 controls.
• Automatic segmentation using QSM and T1-weighted images, with assessment of accuracy and reliability.

Main Results:

• Increased R2 * and QSM observed in portions of all four deep gray matter nuclei in MS subjects compared to controls (P < 0.002).
• Regional analysis with R2 * showed a 66-fold improvement in significance over whole-structure analysis.
• Incorporating QSM improved segmentation accuracy, particularly for the globus pallidus (2.8x higher Dice score), with QSM exhibiting high voxel-based reliability (<1% variation).

Conclusions:

• Quantitative susceptibility mapping (QSM) enhances automatic segmentation of iron-rich deep gray matter.
• Voxel-based evaluation reveals increased R2 * and QSM in all studied deep gray matter nuclei in MS patients.
• Localized R2 * analysis provides superior sensitivity for detecting changes in MS compared to whole-structure measures.