Establishing the boundaries: the hippocampal contribution to imagining scenes

Abstract

When we visualize scenes, either from our own past or invented, we impose a viewpoint for our "mind's eye" and we experience the resulting image as spatially coherent from that viewpoint. The hippocampus has been implicated in this process, but its precise contribution is unknown. We tested a specific hypothesis based on the spatial firing properties of neurons in the hippocampal formation of rats, that this region supports the construction of spatially coherent mental images by representing the locations of the environmental boundaries surrounding our viewpoint. Using functional magnetic resonance imaging, we show that hippocampal activation increases parametrically with the number of enclosing boundaries in the imagined scene. In contrast, hippocampal activity is not modulated by a nonspatial manipulation of scene complexity nor to increasing difficulty of imagining the scenes in general. Our findings identify a specific computational role for the hippocampus in mental imagery and episodic recollection.

Figure 1.

Examples of stimuli used in different conditions. All environments had five structural elements (walls or towers). There were five levels of boundary, from zero to four walls. There were four levels of color, aimed to be of increasing complexity to imagine. No behavioral measures differentiated the levels mixed (1) and mixed (2) so these were combined for the imaging analyses. Subjects were asked to imagine standing on the central yellow dot and visualizing the environments from this location. After forming a clear mental image facing a self-selected direction, they were required to then imagine rotating through 360°.

Figure 2.

Task design. After a variable intertrial interval (ITI), subjects were shown an aerial view of the to-be-imagined scene. The start of the 10 s visualization period was prompted by the instruction to “close your eyes”. An auditory tone signaled the end of the visualization period, and subjects were asked to rate the vividness of their imaged scene using the keypad (rating question). After a response (or after 3.5 s if no response was made), subjects were shown an example screenshot of a ground-level view of a portion of a scene for 1 s and then asked whether it was consistent with their imagined scene (while the picture remained on screen; test question). They had 3.5 s to respond using a four-point confidence rating scale (1, sure it was consistent; 2, fairly sure it was consistent; 3, fairly sure it was inconsistent; 4, sure it was consistent).

Figure 3.

Behavioral results. The relationship between perceived difficulty of imagining the different scenes, the number of boundaries, and color condition.

Figure 4.

Boundary-related activation. There was a parametric modulation of BOLD response by the number of boundaries in four brain regions, including the middle of the left hippocampus. These are shown in the glass brain (A). The percentage signal change in the hippocampus for the five levels of boundary are shown in B. The hippocampal activation is shown projected on to the mean structural image in the top row of C, and the other activations are shown in the bottom row (color bar indicates t-statistic). D shows the percentage signal change associated with the parametric modulation of BOLD by the number of boundaries for the four regions (all trials and trials split by high and low vividness ratings).

Figure 5.

Boundary-related activation during the high vividness trials (color bar indicates t-statistic). The region in the anterior hippocampus/amygdala (left and middle) and the precuneus (right) are shown projected on to the mean structural image.