Perceptual deficits in amnesia: challenging the medial temporal lobe ‘mnemonic’ view

Abstract

Recent animal studies suggest that the medial temporal lobe (MTL), which is thought to subserve memory exclusively, may support non-mnemonic perceptual processes, with the hippocampus and perirhinal cortex contributing to spatial and object perception, respectively. There is, however, no support for this view in humans, with human MTL lesions causing prominent memory deficits in the context of apparently normal perception. We assessed visual discrimination in amnesic cases to reveal that while selective hippocampal damaged patients could discriminate faces, objects, abstract art and colour, they were significantly poorer in discriminating spatial scenes. By contrast, patients with MTL damage, including perirhinal cortex, were significantly impaired in discriminating scenes, faces, and to a lesser extent objects, with relatively intact discrimination of art and colour. These novel observations imply that the human MTL subserves both perceptual and mnemonic functions, with the hippocampus and perirhinal cortex playing distinct roles in spatial and object discrimination, respectively.

Introduction

Multiple studies have demonstrated that activity of hippocampal neurons in rodents reflects aspects of spatial localisation and navigation (O’Keefe, 1976; O’Keefe & Burgess, 1996; O’Keefe, Burgess, Donnett, Jeffery, & Maguire, 1998; Wilson & McNaughton, 1993). Similarly, nonhuman primate data indicate that the hippocampus may only be crucial when spatial memory is critical for a task (Hampton, Hampstead, & Murray, in press; Murray, Davidson, Gaffan, Olton, & Suomi, 1989; Murray, Gaffan, & Mishkin, 1993; Murray & Mishkin, 1998). In contrast, lesion studies have revealed that perirhinal cortex ablations can impair a monkey's ability to make discriminations on the basis of visual feature conjunctions (e.g., object perception), while visual discrimination using single features such as size and colour remains intact (Buckley, Booth, Rolls, & Gaffan, 2001; Bussey et al., 2002, Bussey et al., 2003). Although these findings from the animal literature suggest differential roles for the hippocampus and perirhinal cortex in spatial (Gaffan, 2001, O’Keefe, 1999) and object perception (Buckley, Booth, Rolls, & Gaffan, 2001; Bussey & Saksida, 2002; Bussey et al., 2002, Bussey et al., 2003; Murray & Bussey, 1999), respectively, studies in humans indicate that the same conclusions may not extend to the human medial temporal lobe (MTL). First, while recent research demonstrates some functional homology between rat and human hippocampus, at least in terms of cells that might code for spatial location (Ekstrom et al., 2003), other evidence suggests that the human hippocampus is critical for spatial navigation and memory, rather than for basic spatial perception (Burgess & O’Keefe, 2003). Instead, the adjacent parahippocampal cortex has been highlighted as an important structure for encoding new perceptual information about the layout of spatial scenes (Epstein, Harris, Stanley, & Kanwisher, 1999). Second, patients with perirhinal cortex damage are able to match complex visual stimuli in tasks of minimal mnemonic demand, suggesting normal object perception (Buffalo, Reber, & Squire, 1998; Holdstock, Gutnikov, Gaffan, & Mayes, 2000; Stark & Squire, 2000). This discrepancy between the animal and human literature, however, may reflect the use of paradigms in humans that fail to tax perception of objects that share visual features.

To investigate this possibility, four amnesic patients with selective bilateral hippocampal damage and three amnesic patients with more extensive MTL lesions (Fig. 1; Table 1) were compared to healthy controls on different conditions of two simple computerised visual discrimination tasks. Given that the MTL-lesion patients were significantly older than the hippocampal group, two sets of age-matched controls were recruited. The two discrimination tasks were both based on the same experimental paradigm: On trials 1–3 of each condition, a pair of unfamiliar images from one of five stimulus categories (faces, objects, spatial scenes, abstract art, or colour) was presented on a touchscreen monitor and the subjects were instructed to identify the ‘correct’ picture by touching it. On selection, the correct stimulus produced a high tone, while the incorrect image was associated with a low tone. From trials 4–53, the same pictures were blended together to create 50 new trial unique pairs with five different levels of feature overlap: 0–9%, 10–19%, 20–29%, 30–39%, and 40–49% of shared features. There were 10 trials for each level of blending and these were pseudo-randomly ordered such that two trials from each level were presented during each block of 10 trials. The subjects were instructed to select the picture that they perceived to contain a greater proportion of the original correct stimulus and auditory feedback was provided for all trials. The administration of the different stimulus categories was pseudo-randomised and counterbalanced across all subjects.

The critical difference between the two tasks was that in the first task (Task 1, Fig. 2a) the subjects were required to remember the original target stimulus throughout each experimental condition. In contrast, in the second modified task (Task 2, Fig. 2b) this memory component was removed and the original target item was continually present on the computer screen.