The situation or the person? Individual and task-evoked differences in BOLD activity

Area of Science:

• Neuroimaging
• Cognitive Neuroscience
• Brain Imaging Analysis

Background:

• Recent functional magnetic resonance imaging (fMRI) research shows increased interest in between-person variability.
• Studies report consistent individual differences in blood-oxygenated-level dependent (BOLD) activation and resting-state functional connectivity.
• Conflicting findings exist on whether between-person or within-person cognitive states better explain BOLD activation differences, possibly due to statistical and resolution choices.

Purpose of the Study:

• To investigate the relative contributions of between-person variability and task-condition differences to BOLD activation patterns in fMRI.
• To address discrepancies in previous research by analyzing large-scale task-fMRI data across varying resolutions and distance metrics.

Main Methods:

• Employed multivariate distance matrix regression on multiple task-fMRI datasets.
• Examined variability in BOLD activation patterns at different spatial resolutions, from region-of-interest to vertex-level.
• Utilized unsupervised clustering to identify task structures based on subject-level activation maps.

Main Results:

• Task conditions generally explained more variance in activation map differences than between-person differences at lower resolutions.
• At high-resolution (vertex-level) analyses, between-person differences explained more variance.
• Increasing resolution amplified the contribution of between-person variance while decreasing that of task conditions.
• Clustering revealed a task structure transcending conventional labels, highlighting manipulations that elicit contrasting or similar whole-brain activation.

Conclusions:

• The relative importance of subject versus task effects on BOLD activation is complex and depends on analytical choices.
• Resolution level, distance metric, and task-condition coding significantly influence findings in fMRI variability studies.
• High-resolution analyses may be crucial for capturing genuine between-person differences in brain activation patterns.