Performing group-level functional image analyses based on homologous functional regions mapped in individuals

Area of Science:

• Neuroimaging
• Cognitive Neuroscience
• Human Brain Mapping

Background:

• Traditional functional MRI (fMRI) analysis relies on group-averaged brain atlases, which fail to account for significant intersubject variability in functional organization.
• This limitation hinders accurate functional correspondence and alignment of fMRI data across individuals.

Purpose of the Study:

• To develop and validate a method for identifying discrete, homologous functional regions in individual brains.
• To investigate the impact of these individually defined regions on fMRI data alignment, statistical power, and brain-behavior associations.

Main Methods:

• Utilized resting-state fMRI data to identify homologous functional regions in individual subjects.
• Assessed intersubject variability in the size, position, and connectivity of these regions.
• Compared alignment strategies using individual regions versus group-level atlases for task-fMRI data.
• Examined the relationship between functional connectivity within these regions and fluid intelligence (gF).

Main Results:

• Successfully identified homologous functional regions with significant intersubject variability in size, position, and connectivity.
• Demonstrated that variability in region size and position partially explains previously reported functional connectivity variability.
• Showed that individual differences in network topography correlate with task-evoked activations, validating these regions as individual 'localizers'.
• Aligning task-fMRI data using resting-state derived regions increased statistical power.
• Resting-state connectivity among homologous regions better predicted fluid intelligence than group-atlas derived measures.
• Found that region size and position, in addition to connectivity, relate to human behavior.

Conclusions:

• Identifying homologous functional regions on an individual basis offers a more precise approach to fMRI data analysis.
• This individualized mapping enhances the alignment of task-fMRI data, improving statistical power and the investigation of brain-behavior relationships.
• The findings advocate for incorporating individual functional parcellation in neuroimaging studies for a deeper understanding of brain function and its link to cognition.