Cortical semantization of autobiographical memory over subjective chronological time: An fMRI study

The content and neural representation of autobiographical memories change over time; however, these changes are poorly understood. We hypothesize that the content of memories becomes semanticized, while the neural representation moves from mesial to cortical structures. We conducted an fMRI (functio...

Full description

Saved in:
Bibliographic Details
Published in:The European journal of neuroscience 2022-04, Vol.55 (7), p.1798-1809
Main Authors: McWilliams, Kellie, Kaplan, Alexander, Eline, Ellen, Kaminer, Isabelle, Zodrow, Shauna, Petersen, Julie, Williams, J. Michael, Osipowicz, Karol
Format: Article
Language:English
Subjects:
autobiographical memory
Brain Mapping
Classification
consolidation
Cues
Functional magnetic resonance imaging
Humans
Magnetic Resonance Imaging - methods
Memory
Memory, Episodic
Mental Recall - physiology
multi‐voxel pattern analysis
Neural coding
Temporal Lobe
time
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The content and neural representation of autobiographical memories change over time; however, these changes are poorly understood. We hypothesize that the content of memories becomes semanticized, while the neural representation moves from mesial to cortical structures. We conducted an fMRI (functional magnetic resonance imaging) study on the effects of time on autobiographical memory retrieval. Twenty healthy participants were cued by a selection of photographs that represented distinct episodic memories from 1, 2, 6, and 14 years prior to scanning. Our behavioural data of self‐report measures of memory qualia suggests a loss of episodic content over time. GLM (general linear model) results demonstrate that across all time points, visual association cortices and mesial temporal lobes were activated. However, we did not observe any GLM differences between memory time points. We used SVM (support vector machine) in order to predict memory time point based on neural activation patterns. We were able to accurately predict classification accuracy for the 1‐year (66.7%), 2‐year (66.7%), and 14‐year (33.4%) memory time points, with an overall classification accuracy of 55.6%. We suggest that our findings can be interpreted in light of cortical semantization; as memories age, they become more semanticized and shift in representation towards cortical structures.
ISSN:0953-816X
1460-9568
DOI:10.1111/ejn.15652