Department of Neuroscience and Biomedical Engineering, Aalto University School of Science, Espoo, Finland; BioMag Laboratory, HUS Medical Imaging Center, University of Helsinki and Helsinki University Hospital, Helsinki, Finland.
Department of Neuroscience and Biomedical Engineering, Aalto University School of Science, Espoo, Finland; BioMag Laboratory, HUS Medical Imaging Center, University of Helsinki and Helsinki University Hospital, Helsinki, Finland; Department of Neurology & Stroke and Hertie Institute for Clinical Brain Research, University of Tübingen, Tübingen, Germany.
Neuroimage. 2020 Oct 15;220:117082. doi: 10.1016/j.neuroimage.2020.117082. Epub 2020 Jun 25.
Transcranial magnetic stimulation (TMS) protocols often include a manual search of an optimal location and orientation of the coil or peak stimulating electric field to elicit motor responses in a target muscle. This target search is laborious, and the result is user-dependent. Here, we present a closed-loop search method that utilizes automatic electronic adjustment of the stimulation based on the previous responses. The electronic adjustment is achieved by multi-locus TMS, and the adaptive guiding of the stimulation is based on the principles of Bayesian optimization to minimize the number of stimuli (and time) needed in the search. We compared our target-search method with other methods, such as systematic sampling in a predefined cortical grid. Validation experiments on five healthy volunteers and further offline simulations showed that our adaptively guided search method needs only a relatively small number of stimuli to provide outcomes with good accuracy and precision. The automated method enables fast and user-independent optimization of stimulation parameters in research and clinical applications of TMS.
经颅磁刺激（TMS）方案通常包括手动搜索线圈的最佳位置和方向，或峰值刺激电场，以在目标肌肉中引出运动反应。这种目标搜索非常繁琐，而且结果取决于使用者。在这里，我们提出了一种闭环搜索方法，该方法利用基于先前反应的自动电子刺激调整。电子调整是通过多点 TMS 实现的，刺激的自适应引导是基于贝叶斯优化的原理，以最小化搜索中所需的刺激数量（和时间）。我们将我们的目标搜索方法与其他方法（例如在预定义的皮质网格中进行系统采样）进行了比较。对五名健康志愿者进行的验证实验和进一步的离线模拟表明，我们的自适应引导搜索方法只需要相对较少的刺激就能提供具有良好准确性和精密度的结果。自动化方法能够在 TMS 的研究和临床应用中快速、独立于使用者地优化刺激参数。
