Focusing on what matters: Modulation of the human hippocampus by relational attention

Abstract

Hippocampal episodic memory is fundamentally relational, comprising links between events and the spatiotemporal contexts in which they occurred. Such relations are also important over shorter timescales, during online perception. For example, how do we assess the relative spatial positions of objects, their temporal order, or the relationship between their features? Here, we investigate the role of the hippocampus in online relational processing by manipulating attention to different kinds of relations. While undergoing fMRI, participants viewed two images in rapid succession on each trial and performed one of three relational tasks, judging the images' relative: spatial positions, temporal onsets, or sizes. Additionally, they sometimes judged whether one image was tilted, irrespective of the other. This served as a baseline item task with no demands on relational processing. The hippocampus showed reliable deactivation when participants attended to relational vs. item information. Attention to temporal relations was associated with the most robust deactivation. One interpretation of such deactivation is that it reflects hippocampal disengagement. If true, there should be reduced information content and noisier activity patterns for the temporal vs. other tasks. Instead, multivariate pattern analysis revealed more stable hippocampal representations in the temporal task. This increased pattern similarity was not simply a reflection of lower univariate activity. Thus, the hippocampus differentiates between relational and item processing even during online perception, and its representations of temporal relations are particularly robust. These findings suggest that the relational computations of the hippocampus extend beyond long-term memory, enabling rapid extraction of relational information in perception.

Keywords: attention; medial temporal lobe; relational representations; representational similarity analysis.

FIGURE 1

Experimental design. On every trial, participants were presented with a face and a scene, one above and one below fixation. One image was to the left of the other, one appeared on the screen first, and one was smaller than the other. In addition, each item could independently be tilted clockwise or counterclockwise. Participants were cued before every block of eight trials with the name of the task they were to perform on that block: One of the three possible relational attention tasks (spatial, timing, and size) or the item task. After the presentation of each image pair, participants were shown a response post-cue (either black or gray) that pointed up or down, indicating which item should be used as the reference for the task judgment. Trials lasted 2 s in duration, though the two images were shown on the screen for only 300 ms [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 2

Regions of interest. Example segmentation from one participant is shown for one anterior and one posterior slice. Regions of interest were hand-drawn on individual-participant T2 images. The hippocampal regions of interest were subiculum, CA1, and a combined region of interest for CA2, CA3, and dentate gyrus (CA2/3/ DG). The medial temporal lobe cortex regions of interest were entorhinal cortex [ERC], perirhinal cortex [PRC], and parahippocampal cortex [PHC] [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 3

Univariate evoked activity. Percent signal change for the spatial, timing, and size relational tasks, relative to activity for the item task, in each MTL cortical ROI (a: Parahippocampal cortex [PHC]; b: Perirhinal cortex [PRC]; c: Entorhinal cortex [ERC]) and each hippocampal ROI (d: Subiculum [SUB]; e: CA1; f: CA2/3/DG). Error bars reflect ±1 SEM across subjects [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 4

Multivariate pattern similarity for the timing task. Pearson correlation between activity patterns in each ROI for odd versus even runs of the timing task, relative to the correlation between activity patterns for the item task. Subiculum [SUB] and CA1 showed greater pattern similarity for temporal attention versus item attention. Error bars reflect ±1SEM across subjects. * p < 0.05 [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 5

MUD analysis for the timing task. The contribution of each voxel to pattern similarity was estimated by normalizing BOLD activity over voxels within each ROI, separately for the timing task in even and odd runs, and computing pairwise products across runs. To estimate multivariate-univariate dependence, these products were then correlated with the average univariate activity in the timing task for each voxel. Neither SUB nor CA1 showed a relationship between the two measures, suggesting that deactivated voxels were not solely responsible for increased pattern similarity. Error bars reflect ±1SEM across subjects [Color figure can be viewed at wileyonlinelibrary.com]