Aging and the Role of Prior Knowledge in Item-Level Neural Discrimination of Scene Images

Poster Session F - Tuesday, April 16, 2024, 8:00 – 10:00 am EDT, Sheraton Hall ABC

Yuju Hong¹, Kana Kimura¹, Caitlin Bowman¹; ¹University of Wisconsin-Milwaukee

Using neural pattern information as a window into the contents of memory has become increasingly common in memory research. Aging studies have tended to show age-related dedifferentiation of neural patterns, which may make it difficult to use neural pattern analyses to detect the content of older adults’ memories. Behaviorally, allowing older adults to rely on prior knowledge can help older adults successfully encode new information, but it can also lead to false recognition. How reliance on prior knowledge affects age differences in the discriminability of neural patterns remains unknown. The present study investigated whether prior knowledge of scenes affects neural discrimination of individual scene images during perception and during memory. While undergoing fMRI, both young (18-30 years old) and older (60-80 years old) adults viewed a set of scene images (famous and non-famous manmade and natural landmarks) and then were asked to recall them. To assess neural discriminability, representational similarity analysis (RSA) was performed to assess the strength of item-specific representations compared to same category representations. We performed this analysis separately for famous versus non-famous scenes, as well as in perception versus memory. Preliminary results reveal that the strength of RSA-based item representations were correlated with recognition scores. This effect was present in both young and older adults for both famous and non-famous scenes. Thus, early evidence suggests an age invariant and positive relationship between distinctive neural representations of individual items and behavioral memory discrimination that is present regardless of whether participants have prior knowledge of the memoranda.

Topic Area: LONG-TERM MEMORY: Development & aging