An optimized individual target brain in the Talairach coordinate system.

Area of Science:

• Neuroimaging
• Medical Image Analysis
• Computational Anatomy

Background:

• Spatial normalization aims to align individual brain images to a common template, removing anatomical variations.
• Existing templates, like the single high-resolution brain template (HRBT), can introduce anatomical bias in studies.
• A need exists for optimized templates that better represent group anatomy for accurate brain mapping.

Purpose of the Study:

• To develop and evaluate an optimization method to reduce individual anatomical bias in the ICBM high-resolution brain template (HRBT).
• To create a more representative target brain image for neuroimaging studies by minimizing anatomical discrepancies.
• To improve the accuracy of spatial normalization in three-dimensional brain imaging analysis.

Main Methods:

• The ICBM high-resolution brain template (HRBT) was warped to a group of 27 normal subjects' brain images.
• Displacement fields from these warps were averaged to compute a "minimal deformation target" (MDT) transformation for optimization.
• The optimized HRBT was compared to the original HRBT using spatial cross-correlation metrics.

Main Results:

• Optimization significantly reduced anatomical bias in the HRBT, particularly in specific regions like the superior precentral/postcentral gyri and temporal lobes.
• The optimized HRBT demonstrated statistically significant improvements in anatomical matching to the group of 27 brains (P < 0.05).
• Spatial cross-correlation analysis confirmed enhanced alignment between the group's anatomy and the optimized template.

Conclusions:

• The presented optimization method effectively reduces individual anatomical bias in the HRBT.
• The optimized HRBT provides a more accurate and representative target for spatial normalization in neuroimaging.
• This optimized template facilitates the creation of digital volumetric atlases for normal adult brain anatomy.