BaMS3: Bayesian motor mapping with structured inference for anatomical precision

Area of Science:

• Neuroscience
• Biophysics
• Computational Biology

Background:

• Transcranial magnetic stimulation (TMS) is crucial for non-invasively mapping cortical muscle representations via motor evoked potentials (MEPs).
• Conventional R2-mapping methods independently assess cortical locations, neglecting spatial interactions and nonlinear dynamics, thus limiting precision in complex brain regions.

Purpose of the Study:

• Introduce BaMS3 (Bayesian Motor-Mapping with 3-Stage Structured Fitting), a probabilistic framework for enhanced TMS motor mapping.
• Jointly model spatial sensitivity, nonlinear recruitment dynamics, and variable noise for improved accuracy.

Main Methods:

• BaMS3 integrates anatomical and physiological factors influencing MEP generation.
• Models spatial dependencies, nonlinear input-output relationships, and location-specific variability.
• Validated using synthetic data and empirical datasets from eight healthy participants, compared against R2-based mapping.

Main Results:

• BaMS3 significantly outperforms conventional R2-based mapping in precision and focality.
• Achieved more anatomically precise and spatially focal motor maps.
• Demonstrated greatest improvements in complex cortical regions like sulcal walls and folds.

Conclusions:

• BaMS3 offers a principled probabilistic approach to individualized functional brain mapping.
• Enhances spatial target definition for causal brain mapping, cognitive neuroscience, and preoperative localization.
• Represents a significant advancement in high-resolution functional brain mapping techniques.