The Effects of Attention on Cortical Reinstatement in the Aging Brain

情景记忆的一个决定性特征是对先前经历的事件的生动回忆。情景记忆模型提出，编码过程中出现的相同神经表征在记忆提取过程中也参与其中，这一过程被称为皮层恢复。海马体在稳定记忆痕迹和促进检索方面起着至关重要的作用，通过重新激活编码时存在的感觉特异性皮层区域的相同神经模式。研究表明，早期视觉皮质等感觉特异性皮质区域的神经恢复模式与编码相关海马活动[1]的程度相关，并与较高的记忆准确性[2]相对应。这些发现表明，支持事件编码时复杂表征结合的海马体-皮层连接是“重新体验”先前事件的基础。正常的衰老以情景记忆衰退为特征，这被认为与海马体和周围内侧颞叶结构的年龄相关的结构和功能改变有关。目前的文献集中在年龄相关的记忆衰退是否来自于支持记忆检索的感觉特异性皮层再激活的变化。研究发现，老年人在参与感觉处理的后皮层区域表现出活动减少和特异性下降，并伴有额叶区域活动增加。这种神经模式被认为是由于神经专门化减少而产生的补偿机制。这种领域通用性的神经模式可能导致难以区分存储在记忆中的相似物品。此外，它也可能反映了注意力控制的缺陷，这可能导致前额而不是后感觉区域的过度招募。事实上，老年人通常在行为范式[7]中表现出注意力控制缺陷，额叶区域的过度招募与注意力控制[7]的更大损伤有关。一种可能性是，老年人更依赖于额叶而不是后叶感觉区，这可能反映了他们在引导注意力到感觉信息上的障碍，从而减少了海马将感觉表征与情景记忆联系在一起的程度，以及随后在检索时皮层表征的再激活。我建议研究老年人在记忆提取过程中是否比年轻人更有效地利用感觉处理大脑区域，或者是否注意力控制和编码策略的变化影响了编码感觉信息的本质。为了做到这一点，我建议使用一个任务来操纵记忆编码过程中视觉和语义信息的注意方向。目的是确定是否将注意力转向视觉相对于语义刺激将减少皮层恢复在检索时的年龄差异。为了测量神经活动，我打算使用功能磁共振成像，并特别关注使用表征相似性分析（RSA）[8]。RSA利用表征不相似性矩阵来评估感兴趣区域的神经激活是否与编码和检索条件相似或不同。因此，两种情况下神经模式的相似性越大，表明恢复能力越强，情景记忆的回忆也越好。应用概述的神经成像技术可以比较年轻人和老年人之间的编码检索模式相似性。我预计这项研究的发现将促进我们对衰老大脑中记忆缺陷的神经认知机制的理解。它将通过研究注意力分配是否能促进老年人的皮层恢复效应，以及这与改善对先前编码信息的回忆有何关系，为当前的情景记忆模型做出贡献。

A defining feature of episodic memory is the vivid recollection of previously experienced events. Models of episodic memory propose that the same neural representations present at encoding are engaged during memory retrieval, a process termed cortical reinstatement. The hippocampus plays a crucial role in stabilising memory traces and facilitating retrieval by reactivating the same neural patterns in sensory-specific cortical regions that were present at encoding. Studies show that neural patterns of reinstatement in sensory-specific cortical regions such as early visual cortex are associated with the degree of encoding related hippocampal activity [1] and corresponds with higher memory accuracy [2]. These findings indicate that the hippocampal-cortical connections supporting the binding of complex representations at event encoding underlie the experience of 're-living' a previous event.Normal aging is marked by episodic memory decline, which is thought to be related to age-related structural and functional alterations in the hippocampus and surrounding medial temporal lobe structures [3]. Current literature has focused on whether age-related memory decline emerges from changes in sensory-specific cortical reactivation supporting memory retrieval. Older adults are found to exhibit reduced activity and decreased specificity in posterior cortical regions involved in sensory processing, combined with increased activity in frontal regions [4]. This neural pattern is believed to be a compensatory mechanism resulting from reduced neural specialisation. Such domain-general neural patterns may contribute to the difficulty in distinguishing between similar items stored in memory [5]. Additionally, it could also reflect deficits in attentional control that may lead to an over-recruitment of frontal rather than posterior-sensory regions [6].Indeed, older adults commonly display deficits in attentional control in behavioural paradigms [6] and over-recruitment of frontal regions has been linked to greater impairments in attentional control [7]. One possibility is that the greater reliance on frontal rather than posterior sensory regions in older adults may reflect impairments in directing attention to sensory information, thus reducing the extent to which sensory representations are bound into an episodic memory trace by the hippocampus, and the subsequent reactivation of cortical representations at retrieval.I propose to examine whether older adults recruit sensory processing brain regions during memory retrieval less effectively than younger adults, or if changes in attentional control and encoding strategies affect the nature of encoding sensory information. To do this, I propose to use a task designed to manipulate the orientation of attention of visual and semantic information during memory encoding. The aim is to determine whether orienting attention to visual relative to semantic stimuli will reduce age differences in cortical reinstatement at retrieval.To measure neural activity, I intend to use fMRI with a particular focus on using representational similarity analysis (RSA) [8]. RSA utilises representational dissimilarity matrices to assess whether neural activation in a region of interest is similar or dissimilar for encoding and retrieval conditions. Thus, greater similarity of neural patterns for both conditions indicates greater reinstatement and better episodic memory recollection. Applying the outlined neuroimaging technique allows the comparison of encoding-retrieval pattern similarity between young and older adults.I anticipate that the study findings will advance our understanding of the neurocognitive mechanisms underlying memory deficits in the aging brain. It will contribute to the current models of episodic memory by examining whether attentional allocation can facilitate cortical reinstatement effects in older adults and how this relates to improved recollection for previously encoded information.