Testing for material-specific effects in long-term memory: evidence from amnesia and functional neuroimaging.

A number of diseases, such as epilepsy, dementia, stroke and viral infections, result in memory difficulties, in particular our ability to remember events from the past (long-term memory, LTM). Two regions, the hippocampus and perirhinal cortex, are thought to make an important contribution to LTM as damage to either of these areas results in memory problems. It is unclear, however, whether these two brain regions work together to support LTM, or whether they play distinct roles, particularly for visual memory. The proposed research addresses this question by testing whether the hippocampus is necessary for scene, but not face or object, memory, and if the perirhinal cortex is involved in face and object, but not scene, memory.

Two different techniques are used to address this question. First, we ask patients with damage to the hippocampus and/or perirhinal cortex to perform memory tasks in which they are required to remember whether they have previously seen items from our three visual categories. We wish to know whether damage to these two brain structures results in distinct patterns of memory impairment. Second, we ask healthy participants and patients to undertake memory tasks in an MRI scanner; this allows us to pinpoint which parts of the brain are recruited during memory processing, and to see if the hippocampus and perirhinal cortex show distinct patterns of activity for different visual categories. The findings from these two parts of the grant will be used to refine current ideas about how the hippocampus and perirhinal cortex support LTM, and to develop new models and predictions, that further our knowledge of how damage to these brain areas results in memory difficulties across different diseases.

The final part to the grant aims to use the information we gather from our patient and imaging studies to improve memory in individuals with hippocampal damage. We propose to show that if patients concentrate on the individual details of objects this will help boost their memory for words referring to these objects. If we can demonstrate evidence of transfer from visual to verbal stimuli, this will help clinicians and patients develop better strategies for improving memory in everyday situations.

Long-term memory (LTM) is a critical part of our ability to maintain social interactions and navigate accurately in our environment; not surprisingly, therefore, loss of LTM - common to many neurological disorders (dementia, epilepsy, stroke) - can have a significant impact on independent living and general well being. The high frequency of memory impairment after brain damage also has a major financial cost to society, and a key goal for cognitive and clinical neuroscience therefore, is (a) to understand the genesis of memory impairment, (b) to elucidate the functional role of brain regions considered necessary for memory and (c) to use this scientific knowledge to make progress aimed at intervention after memory impairment.

The proposed project contributes to these important questions by undertaking novel experiments aimed at identifying how two brain regions within the medial temporal lobe (MTL), the hippocampus and perirhinal cortex, contribute to memory for exemplars from different visual categories (faces, objects and scenes). It addresses novel data emerging from patient experiments in which amnesic patients with hippocampal damage do not show a global deficit in LTM (as predicted by most models), but instead reveal memory difficulties that are modulated by category (good memory for faces despite poor memory for scenes, Taylor et al., 2007). Such material-specific differences are not easily explained by current theories of human LTM, and there is an urgent need for more research aimed at obtaining further data to help refine these models.

The experiments proposed here involve neuropsychological and neuroimaging approaches (the latter in both patients and healthy controls) to ask under what circumstances the hippocampus and perirhinal cortex contribute to memory for different visual categories. A unique strength of the proposal is the application of the three main paradigms used in the field to measure LTM (relational processing, memory confidence and source memory); this means that we can test key hypotheses from all existing models, and ensures that we obtain definitive evidence about the role of the hippocampus and perirhinal cortex in visual memory. This knowledge is then used to undertake an experiment aimed at asking whether we can improve verbal memory in amnesia by stressing encoding of object features. This experiment has significant translational potential, and provides a stepping stone to further work aimed at remediating some of day-to-day difficulties experienced by amnesic individuals.