Evaluation of shape-based normalization in the corpus callosum for white matter connectivity analysis

Area of Science:

• Neuroimaging
• Computational Anatomy
• Medical Image Analysis

Background:

• Volumetric registration methods may rely heavily on regularization priors for homogeneous structures.
• Accurate correspondence computation is crucial for analyzing brain structures like the midsagittal section of the corpus callosum (MSCC).

Purpose of the Study:

• To evaluate shape-based normalization using continuous medial representation (cm-rep) for the MSCC.
• To compare cm-rep normalization with traditional volumetric registration paradigms.
• To assess the utility of cm-rep in diffusion tensor tractography (DTT) studies.

Main Methods:

• Utilized continuous medial representation (cm-rep) for shape-based normalization of the MSCC.
• Employed features derived from diffusion tensor tractography (DTT) for validation.
• Compared alignment accuracy and statistical power against volumetric registration.

Main Results:

• Shape-based normalization demonstrated more accurate alignment of MSCC subregions, defined by connectivity, compared to volumetric registration.
• The cm-rep approach improved the statistical power in group analyses involving DTT features.
• This suggests cm-rep is a suitable tool for MSCC normalization in white matter research.

Conclusions:

• Continuous medial representation (cm-rep) provides a robust method for normalizing the midsagittal section of the corpus callosum (MSCC).
• Shape-based normalization enhances the precision of anatomical alignment and statistical analysis in white matter studies using diffusion tensor tractography.