Optimizing Patient-Specific Deep Brain Stimulation Models Using Electrophysiology

项目总结/摘要 脑深部电刺激（DBS）的患者特定计算模型已被用于了解效果 电刺激对帕金森病（PD）和其他神经精神疾病患者大脑结构和通路的影响 紊乱这些三维（3-D），基于图像的，生物物理模型提供了一个视觉表示的神经， DBS电极周围和连接的远端区域中的神经激活模式。因此，他们被控- 在临床实践（靶向和术后编程）和研究（ 机械的见解和假说的发展）。然而，尽管它们在研究中被广泛使用， 最近引入到临床实践中，缺乏对模型准确性的直接评估。目前是 不清楚患者特定计算DBS模型预测的刺激效应的空间范围 反映了人脑中真正的神经元激活。因此，最近的临床研究表明， 简单组织激活体积（VTA）DBS模型预测与一般PD之间的相关性 临床结果。驱动力（DF）预测器DBS模型将更真实的轴突轨迹纳入到神经元模型中。 局部解剖结构;然而，尚不清楚DF模型的额外复杂性是否提高了 模拟。为了验证这一点，我们开发了一种实验范式来量化轴突的程度。 帕金森病患者丘脑底DBS激活神经通路。颅内皮层诱发电位（cEP）和外周- 运动诱发电位（mEP）可以区分DBS对邻近神经通路的激活。 DBS设置（活动触点、振幅、脉冲宽度）的调制会改变cEP和mEP的振幅。 因此，通过改变DBS设置，可以量化通路募集，并且激活预测 不同的建模方法可以进行比较。这项生物工程研究补助金提案的目标是 确定与体内电生理相比，患者特定DBS模型（VTA和DF）的准确性 在PD患者中进行测量。我们假设DF模型比VTA模型在生物病理学上更准确， 埃尔。我们将使用两种不同的电生理测量（Aim 1中的cEP; Aim 2中的mEP）来测试这一点，我们 将确定哪些模型组件对模型模拟最关键。在目标3中，我们将比较 模型预测的神经通路激活与临床DBS副作用（由皮质类固醇引起）之间的关系， 可以通过模型实验测量和预测的NAL/延髓束激活）。这将使我们 以确定个体患者临床使用所需的模型准确度水平。最优 该提案中系统描述的建模方法将为临床提供第一个经过验证的标准， DBS模型的临床和研究应用。

PROJECT SUMMARY/ABSTRACT Patient-specific computational models of deep brain stimulation (DBS) have been used to understand the effects of electrical stimulation on brain structures and pathways in Parkinson's disease (PD) and other neuropsychiatric disorders. These 3-dimensional (3-D), image-based, biophysical models provide a visual representation of neu- ronal activation patterns around the DBS electrode and in connected distal regions. As a result, they are con- ceptually attractive in both clinical practice (for targeting and postoperative programming) and in research (for mechanistic insights and hypotheses development). However, despite their widespread use in research, and recent introduction into clinical practice, the direct assessment of model accuracy is lacking. At present, it is unknown if the spatial extent of stimulation effects predicted by the patient-specific computational DBS models reflect genuine neuronal activations in the human brain. Consequently, recent clinical studies have shown poor correlations between predictions from simple volume of tissue activation (VTA) DBS models and general PD clinical outcomes. Driving force (DF) predictor DBS models incorporate more realistic axonal trajectories into the local anatomy; however, it is unclear if the additional complexity of DF models improves the clinical accuracy of the simulations. To test this, we have developed an experimental paradigm to quantify the degree of axonal pathway activation by subthalamic DBS in PD patients. Intracranial cortical evoked potentials (cEP) and periph- eral motor evoked potentials (mEP) can differentiate activation of several neighboring neural pathways by DBS. Modulation of DBS settings (active contacts, amplitude, pulse width) alters the amplitude of cEP and mEP. Therefore, by changing the DBS settings, pathway recruitment can be quantified, and the activation predictions for different modeling methods can be compared. The goal of this Bioengineering Research Grant proposal is to determine the accuracy of patient-specific DBS models (VTA and DF) compared to in-vivo electrophysiologic measurements in PD patients. We hypothesize that DF models are more biophysically accurate than VTA mod- els. We will test this using two different electrophysiological measurements (cEP in Aim1; mEP in Aim2), and we will identify which model components are the most critical to the model simulations. In Aim 3 we will compare neural pathway activations predicted by the models to clinical DBS side effects (resulting from corticospi- nal/bulbar tract activation that can be experimentally measured and predicted by the models). This will allow us to determine the level of model accuracy that is necessary for clinical use in individual patients. The optimal modeling approach systematically characterized in this proposal will provide the first validated standard for clin- ical and research applications of DBS models.