Consistency and variability in functional localisers.

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Neuroimaging

Background:

• Functional localiser scans (fROIs) are crucial for identifying brain regions of interest in neuroimaging.
• A key assumption is that fROIs identified are consistent across experimental conditions within an individual.
• Intra-subject variability in fROI location can compromise the sensitivity and accuracy of neuroimaging analyses.

Purpose of the Study:

• To investigate the consistency and variability of functional regions-of-interest (fROIs) within individual subjects.
• To assess how different selectivity thresholds affect the reliability of fROI identification.
• To understand the implications of fROI variability for designing functional localiser scans.

Main Methods:

• 45 volunteers underwent two functional localiser scans to identify word- and object-sensitive regions in the occipito-temporal cortex.
• fROIs were defined based on category-selective voxels.
• Spatial overlap between fROIs from the two scans was calculated to measure consistency.

Main Results:

• Voxel consistency was highest using minimally selective thresholds (p<0.05 uncorrected), with approximately 65% overlap between scans.
• Highly selective thresholds (p<10(-4) to 10(-6)) resulted in significantly lower consistency, with less than 25% overlap.
• A substantial proportion of voxels within an fROI (one-third to three-quarters) did not consistently activate across scans.

Conclusions:

• Intra-subject variability in fROIs is surprisingly high, challenging the assumption of stable region identification.
• The choice of statistical threshold significantly impacts the measured consistency of fROIs.
• Findings have critical implications for the design of robust and efficient functional localiser paradigms in neuroimaging research.