BACKGROUND: A key limitation of transcranial magnetic stimulation (TMS) is the lack of real-time, individualized feedback about where the brain is being stimulated, leading to inaccurate assumptions about the underlying affected areas. This uncertainty contributes to poor reproducibility, limited interpretability, and variable efficacy across individuals and sessions. OBJECTIVE: We aimed to develop and validate a neuronavigation system with real-time electric field (E-field) computation and visualization on realistic head models, with extended support for multi-locus TMS (mTMS). METHODS: We designed and built a software framework that integrates an E-field solver with the InVesalius navigation software, introducing a GUI and a workflow that support real-time E-field visualization and quantitative analysis. We characterized our E-field neuronavigation performance and demonstrated it in vivo using a commercial figure-of-eight coil and a robotic-guided 5-coil mTMS. We also quantified precision and accuracy errors, spatial root-mean-square error (RMSE), and maximum absolute gain (MAG). Real-time visualization of electronically shifted E-fields without moving the coil was demonstrated by mapping the precentral gyrus with mTMS and recording motor evoked potentials (MEPs) from finger muscles. RESULTS: The time required for building the boundary element models was 27 s, and the E-field module latency during navigation was approximately 24 ms per coil position. In vivo estimates of E-field distribution during TMS coil repositioning showed low precision errors (0.9-2%) and accuracy errors (2.2-4.9%). MAG values indicated consistent E-field amplitudes relative to reference fields, and RMSE values remained low (< 2.2 V/m) with consistent MEP responses. In mTMS visualization, the MEP amplitude decreased as the electronically shifted E-field moved away from the hotspot. CONCLUSIONS: Our system enables real-time, individualized E-field visualization during TMS, accounting for cortical folding and offering features that simplify, standardize, and improve the reproducibility of the TMS process.
Soto et al. (Wed,) studied this question.