Clinical Connectomics: A network approach to deep brain stimulation

在这个项目中，我们计划-第一次-在不同的空间和时间尺度上结合联合收割机两种计算建模方法-模拟帕金森病（PD）患者的大脑活动，并没有深部脑刺激（DBS）。总体目标是通过这种方法学的发展实现对PD中改变的网络相互作用的更精细和准确的预测，并实现神经调节治疗（如DBS）的计算机模拟规划。通过使用两个不同的模拟器引擎，虚拟脑（TVB）和ANNarchy共同模拟大脑，我们同时捕获全脑网络交互以及基底神经节（BG）网络的详细功能电路交互。这两个描述层次的脑动力学是相互作用和相互影响的。此次合作开发的多尺度模型将由患者特定的多模态数据提供信息，包括脑磁图（MEG），局部场电位（LFPs）和功能性磁共振成像（fMRI）。模型推断的参数和状态变量将用于预测PD患者的临床症状变化。我们将通过交叉验证来验证我们的多尺度模型。开发的软件将是开源的，使其他用户能够使用TVB-ANNarchy工具箱运行多尺度大脑协同模拟。

In this project, we plan – for the first time - to combine two computational modeling approaches at different spatial and temporal scales - to simulate brain activity in Parkinson’s disease (PD) patients with and without deep brain stimulation (DBS). The overarching goal is to achieve with this methodological development more refined and accurate predictions of altered network interactions in PD and to enable in-silico planning of neuromodulation therapies like DBS. By co-simulating the brain with two different simulator engines, The Virtual Brain (TVB) and ANNarchy, we capture whole-brain network interactions as well as detailed functional circuit interactions of the basal ganglia (BG) network simultaneously. Both description levels of brain dynamics are interacting and are mutually influencing each other. The multi-scale models developed in this collaboration will be informed by patient-specific multi-modal data that include magnetencephalography (MEG), local field potentials (LFPs) and functional magnetic resonance imaging (fMRI). Model-inferred parameters and state variables will be used to predict clinical symptom change in PD patients. We will validate our multi-scale models through cross-validation. The developed software will be open-source enabling other users to run multi-scale brain co-simulations with the TVB-ANNarchy toolbox.