Behavior can be decoded across the cortex when individual differences are considered.

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Brain Imaging

Background:

• Traditional group-level analyses link behavior to limited brain regions.
• Individual variability in brain-behavior relationships is often overlooked.

Purpose of the Study:

• To investigate if behavioral metrics like reaction time (RT) and confidence can be decoded across the entire cortex on a subject-specific level.
• To explore the extent and nature of individual differences in brain-behavior mappings.

Main Methods:

• Utilized machine learning models to decode trial-level RT and confidence from brain activation patterns in 50 subjects performing a perceptual decision-making task.
• Split the cortex into 200 regions of interest (ROIs) and analyzed decoding performance per subject and at the group level.
• Developed a robustness test requiring several hundred trials for reliable decoding.

Main Results:

• Group-level decoding confirmed previous findings, identifying RT and confidence in a small percentage of ROIs (12.0% and 3.5%).
• Subject-level decoding revealed that RT and confidence could be decoded from most cortical regions (90.0% and 72.5% respectively) after correction.
• Observed significant individual differences, with many regions showing opposing brain-behavior relationships across subjects.

Conclusions:

• Behavioral signatures are broadly represented across the cortex at the individual level, challenging previous constrained views.
• The study highlights the necessity of considering both group and individual levels to fully understand brain-behavior relationships.
• Findings were validated in a second independent dataset.