Individual-specific resting-state networks predict language dominance in drug-resistant epilepsy

Area of Science:

• Neuroscience
• Medical Imaging
• Epileptology

Background:

• Identifying language dominance is critical for effective epilepsy surgery planning.
• Current methods may be invasive or lack individual specificity.
• Precision functional brain mapping offers a potential solution.

Purpose of the Study:

• To determine if individual-specific cortical network topography can predict language dominance in drug-resistant epilepsy.
• To evaluate a novel precision functional brain mapping approach for this purpose.

Main Methods:

• Developed a multi-session hierarchical Bayesian model (MS-HBM) trained on National Institutes of Health (NIH) data.
• Applied the model to estimate individual-specific resting-state networks in drug-resistant epilepsy patients.
• Validated the model's predictive accuracy for language dominance using neuroimaging data.

Main Results:

• The MS-HBM trained on epilepsy data outperformed group-average networks and generalized well to independent datasets.
• Individual-specific networks showed closer alignment with cortical activity during intracranial electrical stimulation compared to group-average networks.
• Individual-specific language network topography accurately predicted left, bilateral, and right language dominance (AUCs ranging from 0.72 to 0.83).

Conclusions:

• Individual-specific cortical network topography can reliably predict language dominance in drug-resistant epilepsy.
• The developed MS-HBM provides a practical, non-invasive tool for pre-surgical language mapping.
• This approach enhances precision functional brain mapping for epilepsy surgery planning.