Probabilistic atlas for the language network based on precision fMRI data from >800 individuals

Area of Science:

• Neuroimaging
• Cognitive Neuroscience
• Neuroscience

Background:

• Functional magnetic resonance imaging (fMRI) language research traditionally uses group averaging, which is limited by inter-individual variability in language area locations.
• Functional localization, identifying language areas per individual, offers higher sensitivity and interpretability but isn't always feasible or retroactively applicable.

Purpose of the Study:

• To bridge the gap between group-level analysis and individual functional localization in fMRI language research.
• To develop a probabilistic functional atlas of the human language network using a large cohort.

Main Methods:

• Utilized fMRI data from 806 individuals undergoing an extensively validated language localizer task.
• Constructed a probabilistic atlas estimating the likelihood of a location belonging to the language network in a common brain space.

Main Results:

• Developed a probabilistic functional atlas applicable to a common brain space.
• The atlas provides probabilities for any given location belonging to the language network.

Conclusions:

• The probabilistic atlas aids in interpreting group-level activation peaks and lesion locations.
• Facilitates the selection of voxels/electrodes for analysis, enhancing robustness and replicability in fMRI language studies.