Neurocognitive mechanisms of uninstructed dynamic belief updating and pre-clinically relevant interindividual differences in infancy

背景：统计学习中介婴儿期的发展阶段，这是由气质和社会互动塑造。高反应性与以后的焦虑风险有关，而反应性互动促进了社会发展。然而，关于婴儿在不稳定环境下的统计学习以及预测误差（PE）和动态信念更新（DynBU）的神经认知机制尚不清楚。此外，婴儿期DynBU与气质和社会互动经验差异之间的关系，作为潜在的适应不良和临床相关发展的来源，仍未得到探索。婴儿是研究自然附带学习的测试案例。虽然无意识的偶然学习是普遍存在的，但目前即使在成年人中也没有得到很好的理解。目的：我们研究(i)婴儿的神经认知机制与学生在确定性情景下对离散违反期望的反应，以及在预期不确定性下对连续PE的反应；（ii）婴儿对意外不确定性变化点（CP）的DynBU反应，其与皮质网络重置的关系，以及成人非指示偶然DynBU的关系；（iii）与社会互动和气质的临床前维度相关的发展相关的个体间差异。假设：婴儿期的特征可能是对更新的反应相对不足。行为抑制和高反应性可能与过度调节DynBU有关。亲子互动率的降低应该表明确定性经验的减少，并导致DynBU的过度调整，与气质的相互作用。婴儿的初始唤醒应该随着预期不确定性的增加而减少，反应应该对参数PE敏感。CPs会导致觉醒增加，脑电图功率和皮质网络活动发生显著变化。如果未指示的DynBU与明确指示的DynBU具有相似的机制，则成人数据应符合已知的认知模型。计划方法：我们采用RU的任务对1岁婴儿进行偶然视觉预测范式。由于习惯化，或对分布的统计学习，或对分布变化的更新信念，刺激可以成为预期的。我们将测量注视时间、瞳孔变化、erp以及额脑和背脑网络α和θ能量带的振荡。同时，我们将收集婴儿气质和社会互动的行为、神经生理和问卷测量。我们将使用成人的方法为计算模型比较提供数据。预期影响：该项目揭示了DynBU的个体发生起源和神经认知机制，作为进一步发展的起点。它提供了潜在的不适应临床相关发展的个体发生来源的见解。它推进了建模，并为未指示的偶然DynBU提供了生态学上有效的见解。

Background: Statistical learning mediates the developmental phase of infancy, which is shaped by temperament and social interactions. High reactivity is linked to later risks for anxiety, while responsive interactions enhance social development. Still little is known about infants’ statistical learning in volatile environments and the neurocognitive mechanisms of prediction error (PE) and dynamic belief updating (DynBU). Also, the relations between DynBU in infancy and differences in temperament and social-interactional experiences as sources of potentially maladaptive, clinically-relevant development have remained unexplored. Infants are a test case to investigate natural incidental learning. While uninstructed incidental learning is ubiquitous, it is currently not well understood even in adults.Aims: We investigate (i) infants’ neurocognitive mechanisms with pupil responses to discrete violations of expectations in deterministic scenarios, and to continuous PE under expected uncertainty; (ii) infants’ DynBU in response to change points (CP) during unexpected uncertainty, its relation to cortical network reset, and uninstructed incidental DynBU in adults; (iii) developmentally relevant interindividual differences related to social interaction and pre-clinical dimensions of temperament.Hypotheses: Infancy may be characterized by a relative underresponsiveness to updating. Behavioral inhibition and high reactivity should relate to overadjusting DynBU. Reduced rates of parent-infant interactions should indicate decreased experiences of certainty and lead to overadjustment in DynBU, in interaction with temperament. Infants’ initial arousal should decrease with increasingly expected uncertainty, and reactions should be sensitive to parametric PE. CPs should lead to increased arousal and to marked changes in EEG power and cortical network activity. If uninstructed DynBU recruits similar mechanisms as explicit instructed DynBU, adult data should fit the known cognitive models. Planned methods: We adopt the RU’s task to an incidental visual prediction paradigm for 1-year-old infants. Stimuli can become expected either due to habituation, or statistical learning of a distribution, or updated beliefs about changes in the distribution. We will measure looking times, pupillary change, ERPs and oscillations in alpha and theta power bands in frontal and dorsal brain networks. Concurrently, we will collect behavioral, neurophysiological, and questionnaire measures on infants’ temperament and social interaction. We will use the method with adults to provide data for computational model comparisons.Expected impact: The project reveals ontogenetic origins and neurocognitive mechanisms of DynBU as starting point of further development. It provides insights into ontogenetic sources of potentially maladaptive clinically relevant development. It advances modelling and provides ecologically valid insights into uninstructed incidental DynBU.