A tool-box to control and enhance tDCS spatial precision

为了彻底改变我们对大脑的理解，BRAIN倡议呼吁“改进现有的非侵入性神经调节”技术。目前对经颅直流电刺激（tDCS）有很大的兴趣，其是可部署的、耐受性良好的，并且具有靶向神经调节的前景。tDCS的计算模型预测给定电极配置（“蒙太奇”）的个体脑电流，并预测优化的靶向蒙太奇可以实现更多的局灶性皮层刺激。通过三项创新，该提案消除了限制访问计算模型的现有障碍，这些模型将允许研究人员针对所需的皮层目标单独定制电极蒙太奇，以优化临床结果并解决特定的研究假设。首先，在自动图像分割和高通量电流建模方面十年的技术创新将在云支持的开源中得到增强和编码。其次，tDCS电流分布的最新MRI映射将验证和改进模型方法。第三，独立和基于Web的建模客户端软件将部署在服务器上实现的计算要求很高的步骤。只有通过算法优化，建模过程才能分为两个步骤：在服务器上进行基于云的计算密集型处理，然后研究人员在传统PC上使用客户端软件进行模拟。这些创新使得以前需要大量专业知识和劳动力、超级计算机和多次迭代的过程减少到一个步骤，在传统PC上只需几秒钟。此外，我们将提供MRI方案，用于tDCS电流的体内标测。在探索性目的中，MRI标测将测试使用tDCS靶向深部结构的建模预测。直接响应RFA，该提案的结果是优化tDCS空间精度的工具箱，以提高旨在了解大脑和治疗疾病的tDCS研究的严谨性。我们的方法在整合可扩展性、严谨性和透明度方面是独一无二的， 开源（服务器端），具有高度可评估的GUI控制软件（客户端），同时非常强大（例如，非理想的扫描质量）和灵活（例如，传统的垫或高清电极）。

Aimed at revolutionizing our understanding of the brain, the BRAIN initiative calls for “improvement of existing non-invasive neuromodulation” techniques. There is presently great interest in transcranial Direct Current Stimulation (tDCS), which is deployable, well tolerated, and carries the promise of targeted neuromodulation. Computational models of tDCS predict individual brain current flow for a given electrode configuration (“montage”), and predict that optimized targeting montages can achieve more focal cortical stimulation. Through three innovations, this proposal removes existing barriers limiting access to computational models that will allow researchers to individually tailor electrode montages for desired cortical targets so as to optimize clinical outcomes and address specific research hypotheses. First, a decade of technical innovation in automated image segmentation and high- throughput current flow modeling will be enhanced and encoded in cloud-enabled open-source. Second, state-of-the-art MRI mapping of tDCS current distribution will validate and refine model methods. Third, stand-alone and web-based modeling client software will be deployed with computationally demanding steps implemented on servers. Only as a result of algorithmic optimization can the modeling process be divided into two steps: a cloud-based computationally intensive processing on servers, and then simulations taking just seconds by researchers using client software on conventional PC. These innovations result in a process that previously required extensive expertise and labor, super-computers and numerous iterations instead being reduced to a single step, requiring seconds on a conventional PC. In addition, we will supply the MRI protocol for in vivo mapping of tDCS current flow. In an exploratory aim, MRI mapping will test modeling predictions on deep structure targeting with tDCS. Directly responsive to the RFA, the outcome of this proposal is a toolbox for the optimization of tDCS spatial precision to enhance the rigor of tDCS research aimed at understanding the brain and for treating disease. Our approach is unique in integrating the scalability, rigor, and transparency of opens-source (server side) with highly assessable GUI control software (client side), while being exceptionally robust (e.g. non-ideal scan quality) and flexible (e.g. conventional pad or High-Definition electrodes).