Perirhinal contributions to human visual perception

Abstract

Medial temporal lobe (MTL) structures including the hippocampus, entorhinal cortex, and perirhinal cortex are thought to be part of a unitary system dedicated to memory [1, 2], although recent studies suggest that at least one component-perirhinal cortex-might also contribute to perceptual processing [3, 4, 5, 6]. To date, the strongest evidence for this comes from animal lesion studies [7, 8, 9, 10, 11, 12, 13, 14]. In contrast, the findings from human patients with naturally occurring MTL lesions are less clear and suggest a possible functional difference between species [15, 16, 17, 18, 19, 20]. Here, both these issues were addressed with functional neuroimaging in healthy volunteers performing a perceptual discrimination task originally developed for monkeys [7]. This revealed perirhinal activation when the task required the integration of visual features into a view-invariant representation but not when it could be accomplished on the basis of simple features (e.g., color and shape). This activation pattern matched lateral inferotemporal regions classically associated with visual processing but differed from entorhinal cortex associated with memory encoding. The results demonstrate a specific role for the perirhinal cortex in visual perception and establish a functional homology for perirhinal cortex between species, although we propose that in humans, the region contributes to a wider behavioral repertoire including mnemonic, perceptual, and linguistic processes.

Figure 1

Example Stimuli Used in the Experiment, Divided into the Four Main Experimental Conditions Easy (left column) and difficult (right column) trials are shown for object (top rows) and feature (bottom rows) stimuli. It is worth noting that the color reproduction on the page may not exactly correspond to that used in the experiment because of slight differences in the methods used to represent color.

Figure 2

Behavioral Results The top panel displays participants' accuracy scores, and the bottom displays their reaction times for both feature (light gray) and object (dark gray) conditions. In both cases, the difference between easy and difficult trials was highly significant, and for RTs, but not accuracy, there was also a significant interaction. Error bars represent standard error of the mean.

Figure 3

Perirhinal Activation for the Interaction between Stimulus Type and Processing Level The left-hand panels show that the activated voxels (red) within the perirhinal region of interest (yellow) defined as those voxels with a 50% or greater likelihood of being perirhinal cortex from the probabilistic map (Supplement Data). Bar plots in the right panels show the percent regional cerebral blood flow (%rCBF) change for each condition relative to baseline in the right and left perirhinal cortex. Error bars represent standard error of the mean, and significant differences from baseline are indicated with an asterisk (“^∗”). Only the difficult object discriminations led to significant activation relative to baseline in perirhinal cortex. The following abbreviations are used: difficult objects (DO), easy objects (EO), difficult features (DF), easy features (EF).

Figure 4

Entorhinal and Inferotemporal Activation Activation for the interaction between stimulus type and processing level in right entorhinal cortex (A), bilateral posterior fusiform gyrus (B), and right temporopolar cortex (C). In the left column, activations are shown on coronal slices through the group mean structural image with a color scale ranging from Z = 1.66 (red) to Z = 3.0 (yellow). In the right column, bar plots show the percent rCBF relative to baseline per condition for activation in the left (dark gray) and right (light gray) hemispheres. See Figure 3 for abbreviations.