Network-based Near-Scalp Personalized Brain Stimulation Targets

Area of Science:

• Neuroscience
• Medical Imaging
• Computational Biology

Background:

• Functional connectivity (FC) is crucial for personalized transcranial magnetic stimulation (TMS) target identification.
• Current methods often neglect individual whole-cortex network variations.
• Near-scalp targets may enhance patient tolerance by potentially reducing stimulation intensity.

Purpose of the Study:

• Develop an algorithm to optimize both FC and scalp proximity for personalized TMS target localization.
• Improve generalizability across diverse populations by minimizing tunable parameters.

Main Methods:

• Utilized a multi-session hierarchical Bayesian model (MS-HBM) for high-quality, individual-specific cortical network estimation.
• Employed a tree-based algorithm for optimal target location selection.
• Compared the novel approach against 'cluster' and 'cone' algorithms on scalp proximity, reliability, and FC.

Main Results:

• The tree-based MS-HBM reliably identified near-scalp personalized TMS targets for depression in healthy individuals across two datasets.
• Targets demonstrated superior reliability, scalp proximity, and FC to the subgenual anterior cingulate cortex compared to existing methods.
• The algorithm showed versatility by successfully identifying targets for anxiety without parameter tuning.

Conclusions:

• The tree-based MS-HBM algorithm offers a robust and generalizable framework for estimating near-scalp personalized TMS targets.
• This approach has the potential to improve patient tolerance and treatment outcomes in brain stimulation.