Excessive synchronization in neural activity is a hallmark of Parkinson’s disease (PD). A promising technique for treating PD is coordinated reset (CR) neuromodulation in which a neural population is desynchronized by the delivery of spatially-distributed current stimuli using multiple electrodes. In this study, we perform numerical optimization to find the energy-optimal current waveform for desynchronizing neuronal network with CR stimulation, by proposing and applying a new optimization method based on the direct search algorithm. In the proposed optimization method, the stimulating current is described as a Fourier series, and each Fourier coefficient as well as the stimulation period are directly optimized by evaluating the order parameter, which quantifies the synchrony level, from network simulation. This direct optimization scheme has an advantage that arbitrary changes in the dynamical properties of the network can be taken into account in the search process. By harnessing this advantage, we demonstrate the significant influence of externally applied oscillatory inputs and non-random network topology on the efficacy of CR modulation. Our results suggest that the effectiveness of brain stimulation for desynchronization may depend on various factors modulating the dynamics of the target network. We also discuss the possible relevance of the results to the efficacy of the stimulation in PD treatment.

中文翻译：

---

用于使神经元网络动力学不同步的协同复位刺激的数值优化。

神经活动的过度同步是帕金森氏病（PD）的标志。用于治疗PD的一种有前途的技术是协调复位（CR）神经调节，其中通过使用多个电极递送空间分布的电流刺激来使神经群体不同步。在这项研究中，我们通过提出并应用一种基于直接搜索算法的新的优化方法，进行数值优化，以找到使神经元网络与CR刺激不同步的能量最佳电流波形。在所提出的优化方法中，将激励电流描述为傅立叶级数，并且通过评估阶数参数来直接优化每个傅立叶系数以及激励周期，从而从网络仿真中量化了同步水平。这种直接优化方案的优点在于，在搜索过程中可以考虑网络动态特性的任意变化。通过利用这一优势，我们证明了外部应用的振荡输入和非随机网络拓扑对CR调制效果的重大影响。我们的研究结果表明，大脑刺激去同步的有效性可能取决于调节目标网络动力学的各种因素。我们还讨论了结果与PD治疗中刺激效果的可能相关性。我们证明了外部应用的振荡输入和非随机网络拓扑对CR调制效果的重大影响。我们的研究结果表明，大脑刺激去同步的有效性可能取决于调节目标网络动力学的各种因素。我们还讨论了结果与PD治疗中刺激效果的可能相关性。我们证明了外部应用的振荡输入和非随机网络拓扑对CR调制效果的重大影响。我们的研究结果表明，大脑刺激去同步的有效性可能取决于调节目标网络动力学的各种因素。我们还讨论了结果与PD治疗中刺激效果的可能相关性。