Department of Psychology, Renmin University of China, Beijing 100872, China; Laboratory of the Department of Psychology, Renmin University of China, Beijing 100872, China.
School of Artificial Intelligence, Beijing Normal University, Beijing 100093, China.
Neuroscience. 2023 Aug 1;524:158-180. doi: 10.1016/j.neuroscience.2023.05.025. Epub 2023 Jun 5.
The frontoparietal network (FPN) and cingulo-opercular network (CON) may exert top-down regulation corresponding to the central executive system (CES) in working memory (WM); however, contributions and regulatory mechanisms remain unclear. We examined network interaction mechanisms underpinning the CES by depicting CON- and FPN-mediated whole-brain information flow in WM. We used datasets from participants performing verbal and spatial working memory tasks, divided into encoding, maintenance, and probe stages. We used general linear models to obtain task-activated CON and FPN nodes to define regions of interest (ROI); an online meta-analysis defined alternative ROIs for validation. We calculated whole-brain functional connectivity (FC) maps seeded by CON and FPN nodes at each stage using beta sequence analysis. We used Granger causality analysis to obtain the connectivity maps and assess task-level information flow patterns. For verbal working memory, the CON functionally connected positively and negatively to task-dependent and task-independent networks, respectively, at all stages. FPN FC patterns were similar only in the encoding and maintenance stages. The CON elicited stronger task-level outputs. Main effects were: stable CON → FPN, CON → DMN, CON → visual areas, FPN → visual areas, and phonological areas → FPN. The CON and FPN both up-regulated task-dependent and down-regulated task-independent networks during encoding and probing. Task-level output was slightly stronger for the CON. CON → FPN, CON → DMN, visual areas → CON, and visual areas → FPN showed consistent effects. The CON and FPN might together underlie the CES's neural basis and achieve top-down regulation through information interaction with other large-scale functional networks, and the CON may be a higher-level regulatory core in WM.
额顶网络（FPN）和扣带-额顶网络（CON）可能对应于工作记忆（WM）中的中央执行系统（CES）发挥自上而下的调节作用；然而，其贡献和调节机制仍不清楚。我们通过描绘 WM 中 CON 和 FPN 介导的全脑信息流，研究了 CES 下的网络相互作用机制。我们使用参与者执行言语和空间工作记忆任务的数据集，分为编码、维持和探测阶段。我们使用一般线性模型获得任务激活的 CON 和 FPN 节点，以定义感兴趣区域（ROI）；在线元分析定义了验证的替代 ROI。我们使用贝塔序列分析在每个阶段使用 CON 和 FPN 节点计算全脑功能连接（FC）图。我们使用格兰杰因果关系分析获得连接图并评估任务级信息流模式。对于言语工作记忆，CON 在所有阶段分别与任务相关和任务无关网络正相关和负相关地连接。FPN FC 模式仅在编码和维持阶段相似。CON 引起更强的任务级输出。主要效应是：稳定的 CON→FPN、CON→DMN、CON→视觉区域、FPN→视觉区域和语音区域→FPN。在编码和探测过程中，CON 和 FPN 都上调了任务相关网络，下调了任务无关网络。CON 的任务级输出略强。CON→FPN、CON→DMN、视觉区域→CON 和视觉区域→FPN 显示出一致的效果。CON 和 FPN 可能共同构成 CES 的神经基础，并通过与其他大规模功能网络的信息交互实现自上而下的调节，CON 可能是 WM 中更高层次的调节核心。
