Identifying Shared Brain Networks in Individuals by Decoupling Functional and Anatomical Variability.

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Neuroimaging

Background:

• Human brain connectivity architecture exhibits significant inter-individual variability.
• Accurate mapping of functional anatomy at the individual level is crucial for neuroscience and clinical applications.
• Existing methods face challenges in reconciling population-level functional patterns with individual anatomical differences.

Purpose of the Study:

• To develop a novel approach for mapping individual functional brain systems.
• To account for inter-individual variability in functional unit distribution.
• To improve intersubject alignment for enhanced predictability of cognitive functions like language laterality.

Main Methods:

• Utilized resting-state functional connectivity (rsFC) data.
• Developed an "embedding space" to map functional systems based on shared characteristics.
• Employed a clustering algorithm on aligned embedding maps, transforming results back to individual anatomical spaces.
• Compared functional-to-anatomical alignment with purely anatomical alignment.

Main Results:

• Identified reproducible functional systems within subjects, despite varying anatomical locations across individuals.
• Demonstrated improved predictability of individual differences in language laterality using the novel alignment method.
• Revealed that the association strength between function and macroanatomy varies across cortical regions, being stronger in sensorimotor networks and weaker in association networks.

Conclusions:

• The proposed method effectively maps individual functional brain systems while accommodating anatomical variability.
• This approach enhances intersubject alignment, leading to better predictions of individual cognitive differences.
• Findings highlight a gradient in the relationship between brain function and structure across the human cortex.