Representational Approach to Hippocampal Functions : a fMRI Study

Form Perception, Memory, Episodic

memory, neuropsychology, hippocampus, perception, neuroimaging

Although traditional neuropsychological models associate medial temporal lobe (MTL) regions to memory, and the hippocampus to episodic memory, there is growing evidence supporting an alternative view. The representational view of the MTL proposes that such regions depends on the representations processed rather than the cognitive processes. The aim of this project is to characterise the role played by MTL areas, and particularly the hippocampus, in mnemonic and non-mnemonic cognition. Hypotheses stemming from the representational view will thus be tested, using functional MRI among young healthy participants.

Traditional neuropsychological models associate the medial temporal lobe (MTL) with episodic memory. Further, the hippocampus would be responsible for recollection, or the rich and contextualised retrieval of a memory; whereas the perirhinal cortex (PRC) would process familiarity, or the feeling that a stimulus has been encountered before, without remembering where or when it happened. However, there is growing evidence questioning this view, and supporting an alternative proposal. The representational view of the MTL proposes that such regions would not be organised according to cognitive processes, but rather according to the representations (or the mnemonic content, represented by a pattern of neuronal spikes. A 2018-fMRI study demonstrated that the engagement of the hippocampus in recollection depends on the presence of a spatial scene in the memory that is retrieved, and not on the reconstructive nature of the retrieval. However, this demonstration is incomplete, as the only process that was investigated is recollection. The current project aims to extend such results to memory processes such as familiarity, and even to non-mnemonic processes such as visual discrimination. A fMRI study will thus be conducted, where 30 healthy participants will first perform a visual-discrimination task on scenes and objects, before taking a recognition task, in which images will either be presented in full or in the form of circular cues only. Multivariate-Pattern Analyses (MVPA) will complete traditional fMRI analyses, in order to determine whether activity in the MTL can be categorised on the basis of representations through a machine-learning classifier.

No related papers found yet.

France