Neural distinctiveness and discriminability underlying unitization and associative memory in aging

Abstract

Previous work has suggested unitized pairs behave as a single unit and more critically, are processed neurally different than those of associative memories. The current works examines the neural differences between unitization and non-unitized memory using fMRI and multivoxel analyses. Specifically, we examined the differences across face-occupation pairings as a function of whether the pairing was viewed as a person performing the given job (unitized binding) or a person saying they knew someone who had a particular job (non-unitized binding). The results show that at encoding and retrieval, the angular gyrus can discriminate between unitized and non-unitized target trials. Additionally, during encoding, the medial temporal lobe (hippocampus and perirhinal cortex), frontal parietal regions (angular gyrus and medial frontal gyrus) and visual regions (middle occipital cortex) exhibit distinct neural patterns to recollected unitized and non-unitized targets. Furthermore, the perirhinal cortex and medial frontal gyrus show greater neural similarity within subsequently recollected unitized trials compared to non-unitized trials. We conclude that an encoding based strategy to elicit unitization can produce greater associative memory compared to non-unitized trials in older adults. Additionally, when unitized trials are subsequently recollected in the perirhinal cortex older adults show greater neural similarity within unitized trials compared to non-unitized trials.

Keywords: Associative memory; Multivariate pattern analyses; Neural reinstatement; Neuroimaging; Unitization.

Fig. 1

Regions of interest for multivariate classification. Regions defined from AAL pickatlas. Tan = MOC, Dark Red = AG, Blue = MFG, Green = HC, Light Red = PrC. Slice numbers: y = −6; z = 36; x = 32. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Based on Overman and Stephens , Example stimuli for the unitized and associative memory conditions. Face and job pairings were presented on one of two background colors (yellow or blue). Both the encoding conditions contained blue and yellow trials. At retrieval, the pairings were presented on a white background color. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

Based on Haxby et al. , , calculation of neural distinctiveness. Calculations were repeated twice; once for encoding trials and once for retrieval trials. Distinctiveness was calculated by taking the within-condition similarity correlations for all trials within a condition subtracted by the between-condition similarity correlation for all trials between the condition.

Fig. 4

Within-similarity between the two encoding conditions in the AG, HC, MFG, MOC, and PrC between younger and older adults. Neural within-similarity scores significantly different between doing versus speaking conditions indicated with *; p's < 0.05.