Schedule of Events | Symposia

Transentorhinal cortex integrity predicts object fine-grained perceptual and mnemonic discrimination

Poster Session B - Sunday, March 30, 2025, 8:00 – 10:00 am EDT, Back Bay Ballroom/Republic Ballroom

Emma Delhaye^1,2 (emma.delhaye@uliege.be), Gabriel Besson³, Mohamed Ali Bahri², Christine Bastin²; ¹CICPSI, Faculty of Psychology, Lisbon University, ²GIGA-CRC Human Imaging, University of Liège, ³Proaction Lab, CINECC, Faculty of Psychology, University of Coimbra

The transentorhinal cortex (tErC) stands among the first regions to be affected by Alzheimer’s disease (AD) neuropathology, years before any clinical symptoms manifest. Understanding the precise role of this region in cognition is key to detecting the earliest AD-related cognitive impairments. Here, we tested the hypothesis that the tErC supports fine-grained representations of unique individual objects (1) regardless of the cognitive function involved (i.e., perceptual or mnemonic) and, critically, (2) sensitively to the granularity of the demanded discrimination. Therefore, tErC alterations should be consistently and specifically associated with fine-grained object discrimination impairments. We leveraged the prevalence of early AD-related alterations in healthy aging and assessed the association of fine-grained object vs. scene discrimination with two AD-related neuroimaging markers of tErC alterations, i.e., cortical integrity alterations and resting-state functional connectivity changes. Estimating visual similarity using both a pre-trained neural network and subjective measures, our results bring compelling new evidence that the structural integrity of the tErC is specifically related to the sensitivity to visual similarity for objects, but not for scenes, exclusively using neural network similarity measures, but not using human subjective similarity measures, in perceptual discrimination, while in mnemonic discrimination, it is exclusively related with subjective visual similarity ratings but not with similarity computed through a neural network.

Topic Area: LONG-TERM MEMORY: Development & aging