Orientation Prior and Consistent Model Selection Increase Sensitivity of Tract-Based Spatial Statistics in...

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Radiology

Background:

• Diffusion-weighted magnetic resonance imaging (DW-MRI) is crucial for studying white matter microstructure.
• Voxelwise analyses of DW-MRI data, particularly crossing-fiber statistics, face challenges in statistical power and reproducibility.
• Existing frameworks may not fully leverage anatomical information for improved analysis.

Purpose of the Study:

• To enhance the statistical power of crossing-fiber statistics in voxelwise DW-MRI analyses.
• To improve the reproducibility and sensitivity of diffusion models using novel atlas-based approaches.
• To validate the proposed framework against a reference method in a population-based cohort.

Main Methods:

• Developed a framework integrating a fiber orientation atlas and a model complexity atlas.
• Applied the ball-and-sticks model to DW-MRI data from a prospective population-based cohort.
• Utilized tract-based spatial statistics (TBSS) for voxelwise analysis and compared reproducibility (ICC) and sensitivity (age regression) with a reference framework.

Main Results:

• The proposed framework significantly improved the intraclass correlation coefficient (ICC), indicating enhanced reproducibility.
• Higher t-statistics for age regression demonstrated increased sensitivity and precision in detecting biological effects.
• More voxels showed significant correlation with age, suggesting a more robust analysis of white matter changes.

Conclusions:

• Integrating fiber orientation and model complexity atlases significantly enhances the reproducibility and sensitivity of crossing-fiber statistics in TBSS.
• The proposed framework offers a more powerful and precise approach for voxelwise DW-MRI analyses.
• This advancement has implications for population-based studies and understanding age-related white matter changes.