Brain state variability: understanding how the internal world shapes the external world

人们普遍认为，人类大脑在休息期间会继续引发复杂的活动波动，但人们对它们代表什么或它们如何影响行为知之甚少。最近的研究已经开始表明，这些内部波动并不完全是随机的，而是形成了一个准稳定模式的层次结构，现在通常被称为动态大脑状态。然而，目前的研究主要集中在单一模式（如功能磁共振成像（fMRI）），这可能无法完全封装状态的整体。例如，单靠功能磁共振成像无法准确测量睡眠阶段。因此，本项目旨在通过隐马尔可夫模型（HMM）在多模态静息状态数据集中对动态大脑状态进行分类。一旦建立了一组有效可靠的状态，将收集各种感觉任务数据，以探索状态对神经反应的潜在影响。最后，该研究将通过睡眠剥夺研究来控制这些状态。这个项目的结果将提供一组状态，可以作为未来研究大脑状态的框架。此外，如果状态的变化确实在感觉神经反应中占了很大的比例，那么就有可能使用状态信息来帮助“清理”感觉数据集。总体而言，该项目旨在建立一套基于多模态数据的大脑状态，以研究状态对感觉加工的影响。

It is widely accepted that the human brain continues to elicit complex fluctuations in activity during rest, yet little is understood of what they represent or how they influence behaviour. Recent research has begun to show that these internal fluctuations are not solely stochastic, and instead form a hierarchy of quasi-stable patterns, now commonly referred to as dynamic brain states. However, current research has focussed on single modalities (e.g. functional magnetic resonance imaging (fMRI)) which potentially faill to fully encapsulate states in their entirety. For example, fMRI alone cannot accurately measure sleep stages. Therefore, this project aims to classify dynamic brain states via hidden Markov models (HMM) in multimodal resting state data sets. Once a valid and reliable set of states is established, various sensory task data will be collected to explore the potential impact of state on neural responses. Lastly, the study will then aim to manipulate these states via sleep deprivation studies. The results of this project will provide a set of states that may serve as a framework for future investigations into brain state. Furthermore, if variation in state does account for a significant amount of the variance within sensory neural responses, it may be possible to use the state information to help "clean" sensory data sets. Overall, the project aims to establish a set of brain states based on multimodal data to investigate the impact of state on sensory processing.