Emerging fMRI methods quantifying brain dynamics present an opportunity to capture how fluctuations in brain responses give rise to individual variations in affective and motivation states. Although the experience and regulation of affective states affect psychopathology, their underlying time-varying brain responses remain unclear. Here, we present a novel framework to identify network states matched to an affective experience and examine how the dynamic engagement of these network states contributes to this experience. We apply this framework to investigate network state dynamics underlying basal craving, an affective experience with important clinical implications. In a transdiagnostic sample of healthy controls and individuals diagnosed with or at risk for craving-related disorders (total N = 252), we utilized connectome-based predictive modeling (CPM) to identify brain networks predictive of basal craving. An edge-centric timeseries approach was leveraged to quantify the moment-to-moment engagement of the craving-positive and craving-negative subnetworks during independent scan runs. We found that dynamic markers of network engagement, namely more persistence in a craving-positive network state and less dwelling in a craving-negative network state, characterized individuals with higher craving. We replicated the latter results in a separate dataset, incorporating distinct participants (N = 173) and experimental stimuli. The associations between basal craving and network state dynamics were consistently observed even when craving-predictive networks were defined in the replication dataset. These robust findings suggest that network state dynamics underpin individual differences in basal craving. Our framework additionally presents a new avenue to explore how the moment-to-moment engagement of behaviorally meaningful network states supports our affective experiences.

新兴的功能磁共振成像方法可以量化大脑动态，为捕捉大脑反应的波动如何引起情感和动机状态的个体差异提供了机会。尽管情感状态的体验和调节会影响精神病理学，但其潜在的随时间变化的大脑反应仍不清楚。在这里，我们提出了一个新颖的框架来识别与情感体验相匹配的网络状态，并研究这些网络状态的动态参与如何促进这种体验。我们应用这个框架来研究基础渴望背后的网络状态动态，这是一种具有重要临床意义的情感体验。在健康对照和被诊断患有渴望相关疾病或有风险的个体（总N = 252）的跨诊断样本中，我们利用基于连接组的预测模型（CPM）来识别预测基础渴望的大脑网络。利用以边缘为中心的时间序列方法来量化独立扫描运行期间渴望正向和渴望负向子网络的即时参与度。我们发现，网络参与的动态标记，即在渴望积极的网络状态中更持久，在渴望消极的网络状态中更少停留，是具有较高渴望的个体的特征。我们将后一个结果复制到一个单独的数据集中，其中包含不同的参与者（ N = 173）和实验刺激。即使在复制数据集中定义了渴望预测网络，基本渴望和网络状态动态之间的关联仍然被观察到。这些强有力的发现表明，网络状态动态支撑了基础渴望的个体差异。 我们的框架还提供了一个新的途径来探索具有行为意义的网络状态的即时参与如何支持我们的情感体验。