Corpus callosum subdivision based on a probabilistic model of inter-hemispheric connectivity

Area of Science:

• Neuroimaging
• Statistical Shape Analysis
• Computational Anatomy

Background:

• The corpus callosum (CC) is crucial for interhemispheric communication.
• Accurate CC subdivision is essential for studying its morphology and development.
• Existing CC subdivision methods lack reproducibility and stability.

Purpose of the Study:

• To develop a novel, automatic, and reproducible method for probabilistic CC subdivision.
• To combine statistical shape analysis with Diffusion Tensor Image (DTI) tractography.
• To assess regional CC area growth in young children.

Main Methods:

• Utilized Diffusion Tensor Image (DTI) tractography to identify trans-callosal fibers.
• Employed statistical shape analysis on CC contour points.
• Developed a probabilistic subdivision model based on fiber distances and automatic lobe subdivision.
• Applied the method to a cohort of healthy subjects aged 2-4 years.

Main Results:

• The proposed method provides a probabilistic, model-based CC subdivision.
• Results demonstrated higher stability and reproducibility compared to Witelson and bounding box methods.
• Successfully applied to analyze regional CC area growth in 2-4 year olds.

Conclusions:

• The novel combination of shape analysis and DTI tractography offers a robust CC subdivision technique.
• This method enhances the study of CC morphology and developmental changes.
• The approach is automatic, reproducible, and suitable for pediatric neuroimaging research.