Neural Network Bases of Healthy and Aberrant Belief Updating

背景：动态信念更新是健康认知的关键，精神分裂症患者的动态信念更新似乎受到损害。最近的进展为动态信念更新背后的大规模脑回路操作提供了新的见解。大脑皮质专门网络中信念更新计算的主动重置，由脑干觉醒系统的瞬时(相)激活触发，可能在这一过程中发挥关键作用。临床研究发现，精神分裂症患者的基线觉醒水平存在异常。我们认为，基线觉醒的异常减少了信念更新过程中的相位性唤醒反应；这损害了在环境变化后更新信念状态所需的大规模皮层网络重置。我们计划解开健康人脑中这一假定的机械级联反应，并量化其在精神分裂症中的异常。目标与假设：我们的目标是：(I)识别大规模信号，并破译皮层网络重置在变化点之后的功能作用；(Ii)揭示任务诱发的相位性唤醒反应、紧随变化点的唤醒反应对基线唤醒水平的依赖；以及(Iii)量化精神分裂症的相位性唤醒和皮质网络重置的像差。根据试验结果，我们预计改变点触发的皮质网络重置将在证据编码与信念编码的皮质区域的神经可变性的不同调制中表现出来。这将促进一种新的信仰状态取代旧的信仰状态。我们进一步预计，基线唤醒水平较高的健康个体将表现出更少的变化点触发的相位性唤醒反应，这将转化为更低效率的网络重置和更不灵活的信念更新。最后，我们预计这种基线唤醒依赖的信念更新级联损伤可能在精神分裂症中被放大。方法：我们计划综合应用行为建模、通过经验采样在日常生活中的唤醒跟踪、通过斜视测量和心电图对唤醒的实验室评估以及大规模皮层网络动力学的脑磁图(MEG)评估来验证这些假说。新颖的分析技术将使我们能够以高度的空间和时间特异性询问人类大脑中潜在的信念更新的大规模电路操作。为了为我们的结论奠定坚实的基础，我们将评估来自健康大脑的多个现有的瞳孔测量-脑磁图数据集中的网络重置信号。在一个新的数据集中，我们将比较精神分裂症患者和健康对照组之间的觉醒和皮质信念更新动力学。影响：该项目将推进对健康和异常信念更新的基本理解，并可能为精神分裂症患者脑干唤醒系统和皮质计算的相互作用的生物标志物分析铺平道路。

Background: Dynamic belief updating is key for healthy cognition and seems to be impaired in schizophrenia. Recent advances have yielded new insights into the large-scale brain circuit operations underlying dynamic belief updating. The active reset of belief updating computations in specialized networks of the cerebral cortex, triggered by the transient (phasic) activation of brainstem arousal systems, may play a key role in this process. Clinical research has discovered aberrations of baseline arousal levels in schizophrenia. We propose that aberrations of baseline arousal reduce phasic arousal responses during belief updating; this impairs the large-scale cortical network resets required for updating the belief state following environmental changes. We plan to unravel this putative mechanistic cascade in the healthy human brain and quantify its aberrations in schizophrenia. Aims & hypotheses: We aim to (i) identify large-scale signatures and decipher the functional role of cortical network resets following change points; (ii) unravel the dependence of task-evoked, phasic arousal responses, following change points, on baseline arousal levels; and (iii) quantify aberrations of phasic arousal and cortical network reset in schizophrenia. Based on pilot results, we expect that change point-triggered cortical network resets will manifest in distinct modulations of neural variability in evidence-encoding versus belief-encoding cortical areas. This will facilitate the substitution of an old by a new belief state. We further expect that healthy individuals with higher baseline arousal will exhibit reduced change point-triggered phasic arousal responses, which will translate into less effective network resets and less flexible belief updating. Finally, we expect that such baseline arousal-dependent impairments of the belief-updating cascade may be amplified in schizophrenia. Methods: We plan to test these hypotheses with an integrated application of behavioral modeling, arousal tracking in daily life through experience sampling, laboratory assessment of arousal through pupillometry and ECG, and magnetoencelphalographic (MEG) assessment of large-scale cortical network dynamics. Novel analysis techniques will enable us to interrogate the large-scale circuit operations underlying belief updating in the human brain with high spatial and temporal specificity. To build a solid foundation for our conclusions, we will assess network reset signatures in multiple existing pupillometry-MEG data sets from the healthy brain. In a new data set, we will compare arousal and cortical belief updating dynamics between schizophrenia patients and healthy controls. Impact: This project will advance the fundamental understanding of healthy and aberrant belief updating and may pave the way for a biomarker assay of the interplay of brainstem arousal systems and cortical computation in schizophrenia.