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

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

Kana Kimura¹ (kkimura@uwm.edu), Yuju Hong¹, Caitlin R. Bowman¹; ¹University of Wisconsin - Milwaukee

It is well known that the detail and specificity of episodic memory declines with older age. Prior work has shown that using semantic knowledge can help older adults successfully encode new information but can also lead to false recognition. Using neural pattern information as a window into the contents of memory has become increasingly common, and these analyses allow researchers to assess the neural patterns that are consistently evoked by items in the same category. Whether having prior knowledge of to-be-learned stimuli increases processing of category-level information in older adults remains unknown. In the present study, both young (18-30 years old) and older adults (60-80 years old) viewed a set of scene images (mix of famous and non-famous manmade and natural landmarks) and then were asked to recall them all while undergoing fMRI. To assess neural representations of category-level information, we trained and tested a multivariate classifier to distinguish manmade from natural scenes separately based only on the famous locations or only the non-famous locations. We also did this separately for perception and memory recall. Preliminary results revealed an age deficit in remembering non-famous scenes with comparable performance for famous scenes. We observed better decoding of scene category for young adults in perception, while older adults had an advantage in memory recall. These effects emerged for both famous and non-famous scenes. Thus, age differences in category representations may be driven more strongly by task than by prior familiarity with memoranda.

Topic Area: LONG-TERM MEMORY: Development & aging