An epileptogenic focus in the dominant temporal lobe can result in the reorganization of language systems in order to compensate for compromised functions. We studied the compensatory reorganization of language in the setting of left temporal lobe epilepsy (TLE), taking into account the interaction of language (L) with key non-language (NL) networks such as dorsal attention (DAN), fronto-parietal (FPN) and cingulo-opercular (COpN), with these systems providing cognitive resources helpful for successful language performance.

We applied tools from dynamic network neuroscience to functional MRI data collected from 23 TLE patients and 23 matched healthy controls during the resting state (RS) and a sentence completion (SC) task to capture how the functional architecture of a language network dynamically changes and interacts with NL systems in these two contexts.

We provided evidence that the brain areas in which core language functions reside dynamically interact with non-language functional networks to carry out linguistic functions. We demonstrated that abnormal integrations between the language and DAN existed in TLE, and were present both in tonic as well as phasic states. This integration was considered to reflect the entrainment of visual attention systems to the systems dedicated to lexical semantic processing. Our data made clear that the level of baseline integrations between the language subsystems and certain NL systems (e.g., DAN, FPN) had a crucial influence on the general level of task integrations between L/NL systems, with this a normative finding not unique to epilepsy. We also revealed that a broad set of task L/NL integrations in TLE are predictive of language competency, indicating that these integrations are compensatory for patients with lower overall language skills.

We concluded that RS establishes the broad set of L/NL integrations available and primed for use during task, but that the actual use of those interactions in the setting of TLE depended on the level of language skill. We believe our analyses are the first to capture the potential compensatory role played by dynamic network reconfigurations between multiple brain systems during performance of a complex language task, in addition to testing for characteristics in both the phasic/task and tonic/resting state that are necessary to achieve language competency in the setting of temporal lobe pathology. Our analyses highlighted the intra- versus inter-system communications that form the basis of unique language processing in TLE, pointing to the dynamic reconfigurations that provided the broad multi-system support needed to maintain language skill and competency.

主要颞叶的癫痫病灶可导致语言系统的重组，以补偿受损的功能。我们研究了左颞叶癫痫（TLE）背景下语言的代偿性重组，考虑到语言（L）与关键非语言（NL）网络（如背侧注意（DAN）、额顶叶（））的相互作用。 FPN）和扣带盖（COpN），这些系统提供有助于成功的语言表现的认知资源。

我们将动态网络神经科学的工具应用于从 23 名 TLE 患者和 23 名匹配的健康对照在静息状态 (RS) 和句子完成 (SC) 任务中收集的功能性 MRI 数据，以捕捉语言网络的功能架构如何动态变化和交互在这两种情况下使用 NL 系统。

我们提供的证据表明，核心语言功能所在的大脑区域与非语言功能网络动态交互以执行语言功能。我们证明了语言和 DAN 之间的异常整合存在于 TLE 中，并且存在于主音和相位状态中。这种整合被认为反映了视觉注意系统对专门用于词汇语义处理的系统的夹带。我们的数据清楚地表明，语言子系统和某些 NL 系统（例如，DAN、FPN）之间的基线集成水平对 L/NL 系统之间的任务集成的一般水平具有至关重要的影响，这不是唯一的规范性发现癫痫。我们还揭示了 TLE 中广泛的任务 L/NL 集成可以预测语言能力，

我们得出的结论是，RS 建立了广泛的 L/NL 集成集，并准备在任务期间使用，但在 TLE 设置中这些交互的实际使用取决于语言技能水平。我们相信我们的分析是第一个捕捉到在执行复杂语言任务期间多个大脑系统之间的动态网络重新配置所发挥的潜在补偿作用，此外还测试了必要的阶段/任务和强直/休息状态的特征在颞叶病理学环境中实现语言能力。我们的分析强调了构成 TLE 中独特语言处理基础的系统内和系统间通信，