Automated mapping with multi-locus TMS

The multi-locus TMS technology [1] allows us to electronically shift the target location of a TMS pulse. Thus, we can precisely map stimulated areas to responses. For example, one can systemically search for the target where the strongest motor response occurs. This motor response can be measured using electromyogram (EMG). Searching for the optimal target location is usually done by hand prior to an experiment. Manually, this can be a time-consuming process requiring a skilled researcher. We aim to speed up these processes to automize them in multi-locus TMS using machine learning methods. Systematically testing each possible location would be a very lengthy process but we can use algorithms like Bayesian Optimization to avoid this. These can calculate what areas are likely to improve our outcome and only test these. Thus, we are improving speed and reliability of our result. By also using prior expert knowledge, we can combine the benefits of searching by hand and using machine learning. This allows a fast and precise mapping of the mTMS target area. For the development of these methods, we use simulations and experiments.

[1] Nieminen, Jaakko O., et al. "Multi-locus transcranial magnetic stimulation system for electronically targeted brain stimulation." Brain stimulation 15.1 (2022): 116-124.

Project contributors: Miriam Kirchhoff

Funding: European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (ConnectToBrain, ERC synergy grant agreement No. 810377)