On testing for spatial correspondence between maps of human brain structure and function.

Area of Science:

• Neuroimaging
• Neuroscience
• Brain Mapping

Background:

• Comparing brain maps across studies is crucial but lacks standardized methods.
• The "correspondence problem" hinders interpretation of functional activation, resting-state networks, and anatomical landmarks.
• Existing methods for comparing brain maps vary significantly in rigor and approach.

Purpose of the Study:

• To address the "correspondence problem" in neuroimaging by developing a statistically grounded method for comparing brain maps.
• To provide a robust framework for testing hypotheses about the spatial overlap between different types of brain data.
• To facilitate the integration of diverse neuroimaging datasets.

Main Methods:

• Developed a spatial permutation framework using random rotations on spherical cortical representations.
• Provided a theoretical statistical foundation for the proposed spatial permutation approach.
• Applied the method to derive clusters of cognitive functions based on their neuroanatomical substrates.

Main Results:

• Demonstrated the correspondence between task-based functional activity, resting-state functional magnetic resonance imaging (fMRI) networks, and gyral-based anatomical landmarks using public data.
• Derived clusters of cognitive functions correlated with specific functional neuroanatomical substrates.
• The spatial permutation approach offers a statistically rigorous method for comparing cortical maps.

Conclusions:

• The spatial permutation framework provides a significant advancement for comparing cortical maps.
• This method enhances the integration of diverse neuroimaging data, enabling new research possibilities.
• Open-access code is available to implement these methods for surface-based cortical analysis.