The ability to sustain attention differs across people and varies over time within a person. Models based on patterns of static functional brain connectivity observed during task performance and rest show promise for predicting individual differences in sustained attention as well as other forms of attention. The sensitivity of connectome-based models to attentional state changes, however, is less well characterized. Here, we review recent evidence that time-varying functional brain connectivity predicts fluctuations in attention in controlled and naturalistic task contexts. We propose that building connectome-based models to predict changes in attention across multiple timescales and experimental contexts can help further disentangle state versus trait influences on functional connectivity patterns, elucidate the behavioral relevance of functional connectivity dynamics, and contribute to the development of a comprehensive suite of generalizable neuromarkers of attention. To achieve this goal, we suggest collecting multi-task, multi-session neuroimaging samples with concurrent behavioral and physiological measures of attentional state.

维持注意力的能力因人而异，并且随着时间的推移而变化。基于在任务执行和休息期间观察到的静态功能性大脑连接模式的模型显示出有望预测持续注意力以及其他形式注意力的个体差异。但是，基于连接组的模型对注意力状态变化的敏感性较差。在这里，我们回顾了最近的证据，即时变的功能性大脑连通性可预测受控和自然主义任务环境中注意力的波动。我们建议，建立基于连接组的模型来预测跨多个时间尺度和实验环境的注意力变化，可以帮助进一步消除状态和特征对功能连接模式的影响，阐明功能连接动力学的行为相关性，并有助于开发一套综合的可关注的通用神经标记。为实现此目标，我们建议收集多任务，多会话的神经影像样本，同时注意注意力状态的行为和生理指标。