A spline-based regression parameter set for creating customized DARTEL MRI brain templates from infancy to old age

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biostatistics

Background:

• Brain MRI analysis is crucial for understanding neurological development and aging.
• Accurate population-specific templates are needed for consistent neuroimaging studies.
• Existing methods may not capture age-related trajectories effectively.

Purpose of the Study:

• To derive regression parameters from segmented brain MRI data.
• To develop a method for generating age-and-demographic-matched brain templates.
• To provide a publicly available dataset and algorithm for neuroimaging research.

Main Methods:

• Analysis of 1919 brain MRI datasets (gray and white matter) from healthy subjects (ages 1-75).
• Image segmentation using CAT12 and affine-only spatial normalization.
• Iterative non-linear registration (DARTEL procedure) for image crispness.
• Multivariate adaptive regression splines (MARS) analysis via CerebroMatic toolbox.

Main Results:

• Regression parameters characterizing gray and white matter changes across the lifespan.
• Successful modeling of smoothly varying trajectories considering age, gender, and technical factors.
• Generation of parameters for creating matched DARTEL or SHOOT templates.

Conclusions:

• The derived regression parameters facilitate the creation of customized brain templates for diverse populations.
• This resource supports neuroimaging research from infancy to old age.
• The dataset and algorithm are publicly accessible, promoting reproducibility and further research.