Transcranial magnetic stimulation (TMS)-based motor mapping is an established method for non-invasively localizing cortical muscle representations from motor evoked potentials (MEPs). Conventional vertex-wise mapping approaches treat cortical locations independently and do not account for spatial interactions, non-linear recruitment dynamics, or variable trial-to-trial noise. This can limit spatial precision, particularly in anatomically complex and folded regions of the cortex.
In their new study, David Luis Schultheiss, Zsolt Turi, Andreas Vlachos and Joschka Boedecker introduce BaMS3 (Bayesian motor mapping with 3-stage structured fitting), a probabilistic framework for TMS-based motor mapping. BaMS3 jointly models spatial sensitivity, non-linear recruitment dynamics and variable noise. The framework incorporates anatomical and physiological determinants of motor evoked potentials by modeling spatial dependencies across cortical locations, non-linear input–output relationships and location-specific variability.
The authors evaluated BaMS3 using subject-specific synthetic simulations and empirical datasets from eight healthy participants. The method was compared with conventional R²-based mapping with regard to hotspot localization accuracy and motor map focality.
The results show that BaMS3 produced more anatomically precise and spatially focal motor maps while preserving canonical hotspot localization in the empirical datasets. The largest improvements were observed in anatomically complex cortical regions, including sulcal walls and cortical folds.
By modeling TMS motor mapping as a probabilistic inference problem, BaMS3 provides a framework for individualized functional brain mapping and may improve spatial target definition for causal brain mapping, cognitive neuroscience and preoperative functional localization.
Publication:
Schultheiss, D. L., Turi, Z., Vlachos, A., & Boedecker, J. (2026). BaMS3: Bayesian motor mapping with structured inference for anatomical precision. Journal of Neural Engineering, 23(4), 046051.
DOI: 10.1088/1741-2552/ae8eb2
Read the full publication here.
