📄 中文摘要
控制癫痫发作期间高维神经动态仍然是一个重大挑战，主要由于大脑的非线性特征和复杂的连接性。提出了一种新颖的框架，即图正则化库曼均场博弈（GK-MFG），该框架结合了用于库曼算子近似的水库计算（RC）和用于解决分布控制问题的交替种群与代理控制网络（APAC-Net）。通过将脑电图（EEG）动态嵌入线性潜在空间，并施加基于相位锁定值（PLV）导出的图拉普拉斯约束，该方法在尊重大脑功能拓扑结构的同时，实现了稳健的癫痫抑制。
驯服癫痫：全脑动态的均场控制
出处: Taming Epilepsy: Mean Field Control of Whole-Brain Dynamics
发布: 2026年3月20日
📄 English Summary
Taming Epilepsy: Mean Field Control of Whole-Brain Dynamics
Controlling high-dimensional neural dynamics during epileptic seizures poses a significant challenge due to the brain's nonlinear characteristics and complex connectivity. A novel framework, named Graph-Regularized Koopman Mean-Field Game (GK-MFG), is proposed, integrating Reservoir Computing (RC) for Koopman operator approximation with Alternating Population and Agent Control Network (APAC-Net) for addressing distributional control problems. By embedding Electroencephalogram (EEG) dynamics into a linear latent space and imposing graph Laplacian constraints derived from the Phase Locking Value (PLV), this method achieves robust seizure suppression while respecting the functional topological structure of the brain.