Measuring brain variability via sulcal lines registration: a diffeomorphic approach

Area of Science:

• Neuroimaging
• Computational Anatomy
• Medical Image Analysis

Background:

• Quantifying brain variability is crucial for understanding neurological development and disease.
• Existing methods for measuring brain variability, such as tensor extrapolation, have limitations in defining variability consistently across the entire brain space.
• Sulcal lines offer rich anatomical information for characterizing brain morphology.

Purpose of the Study:

• To develop a new method for measuring brain variability using sulcal line registration.
• To overcome limitations of previous methods requiring tensor extrapolation.
• To provide a more comprehensive and globally consistent measure of brain variability.

Main Methods:

• Modeling sulcal lines geometrically as currents to avoid point correspondence matching.
• Employing a large deformation framework for registration of sulcal line sets.
• Utilizing Tangent-Principal Component Analysis (Tangent-PCA) for analyzing deformation trends.

Main Results:

• The new method provides a globally consistent deformation of the brain space.
• Variability is defined everywhere, unlike previous methods requiring extrapolation.
• The approach reveals different details of variability and highlights major deformation trends in a database.

Conclusions:

• The proposed method offers an improved approach to measuring brain variability.
• Geometric modeling of sulcal lines as currents enhances registration accuracy and consistency.
• This framework facilitates a more detailed and comprehensive understanding of brain deformation patterns.