Lu Shujie, Jiang Haoyu, Li Chengwei, Hong Baoyu, Zhang Pu, Liu Wenli
Center for Medical Metrology, National Institute of Metrology, Beijing, China.
China Academy of Telecommunications Technology, Beijing, China.
Front Public Health. 2022 Mar 11;9:794167. doi: 10.3389/fpubh.2021.794167. eCollection 2021.
Transcranial magnetic stimulation (TMS), a non-invasive technique to stimulate human brain, has been widely used in stroke treatment for its capability of regulating synaptic plasticity and promoting cortical functional reconstruction. As shown in previous studies, the high electric field (E-field) intensity around the lesion helps in the recovery of brain function, thus the spatial location and angle of coil truly matter for the significant correlation with therapeutic effect of TMS. But, the error caused by coil placement in current clinical setting is still non-negligible and a more precise coil positioning method needs to be proposed. In this study, two kinds of real brain stroke models of ischemic stroke and hemorrhagic stroke were established by inserting relative lesions into three human head models. A coil position optimization algorithm, based on the genetic algorithm (GA), was developed to search the spatial location and rotation angle of the coil in four 4 × 4 cm search domains around the lesion. It maximized the average intensity of the E-field in the voxel of interest (VOI). In this way, maximum 17.48% higher E-field intensity than that of clinical TMS stimulation was obtained. Besides, our method also shows the potential to avoid unnecessary exposure to the non-target regions. The proposed algorithm was verified to provide an optimal position after nine iterations and displayed good robustness for coil location optimization between different stroke models. To conclude, the optimized spatial location and rotation angle of the coil for TMS stroke treatment could be obtained through our algorithm, reducing the intensity and duration of human electromagnetic exposure and presenting a significant therapeutic potential of TMS for stroke.
经颅磁刺激（TMS）是一种用于刺激人脑的非侵入性技术，因其具有调节突触可塑性和促进皮质功能重建的能力，已被广泛应用于中风治疗。如先前研究所示，病变周围的高电场（E场）强度有助于脑功能恢复，因此线圈的空间位置和角度对于与TMS治疗效果的显著相关性至关重要。但是，当前临床环境中由线圈放置引起的误差仍然不可忽略，需要提出一种更精确的线圈定位方法。在本研究中，通过在三个人头模型中插入相关病变，建立了缺血性中风和出血性中风两种真实脑中风模型。开发了一种基于遗传算法（GA）的线圈位置优化算法，以在病变周围的四个4×4 cm搜索域中搜索线圈的空间位置和旋转角度。它使感兴趣体素（VOI）中的E场平均强度最大化。通过这种方式，获得的E场强度比临床TMS刺激的E场强度最高高17.48%。此外，我们的方法还显示出避免不必要地暴露于非目标区域的潜力。所提出的算法经过九次迭代验证可提供最佳位置，并且在不同中风模型之间的线圈位置优化方面显示出良好的鲁棒性。总之，通过我们的算法可以获得用于TMS中风治疗的线圈优化空间位置和旋转角度，减少人体电磁暴露的强度和持续时间，并呈现出TMS对中风显著的治疗潜力。
