Age-related changes in the neural representation of naturalistic events

1194 

Abstract Submission 

Selma Lugtmeijer¹, Djamari Oetringer², Linda Geerligs³, Karen Campbell⁴

¹University of Birmingham, Birmingham, West Midlands, ²Radboud University, Nijmegen, Gelderland, ³Donders Institute, Nijmegen, Gelderland, ⁴Brock Univeristy, St. Catharines, Ontario

Selma Lugtmeijer  
University of Birmingham
Birmingham, West Midlands

Djamari Oetringer  
Radboud University
Nijmegen, Gelderland

Linda Geerligs  
Donders Institute
Nijmegen, Gelderland

Karen Campbell  
Brock Univeristy
St. Catharines, Ontario

How do people segment ongoing experiences into separate events? Segmentation is a crucial process that improves understanding of current events (Zacks et al., 2001) and the recall of past events (Flores et al., 2017). Naturalistic stimuli like movies provide an opportunity to study event segmentation with functional magnetic resonance imaging (fMRI). Transitions between distinct neural states may be the mechanism underlying event segmentation (Baldassano et al., 2017). Neural state boundaries are organized in a hierarchical manner across the cortex, with short states in primary sensory regions, and long states in the prefrontal cortex (Geerligs, 2022). Perceived event boundaries overlap with neural state changes across the cortical hierarchy, especially state changes that are shared between many regions. An open question is whether these neural state changes related to event segmentation are stable across the adult lifespan.

Participants from a lifespan cohort (N = 577, age 18–88; Cam-CAN; Shafto et al., 2014) viewed an 8 min movie during an fMRI scan. To identify neural state boundaries, we applied the greedy state boundary search (GSBS; Geerligs et al., 2021). Participants were sorted into 34 age groups (Geerligs et al., 2021). GSBS identifies the optimal number of state boundaries based on brain activity time courses from spherical searchlights covering the entire cortex. This metric identifies the optimal number of state boundaries within a timeseries, such that the correlations of timepoints within a state are maximized and correlations of timepoints in consecutive states are minimized. Perceived event boundaries were based on an external dataset in which participants were asked to indicate when they felt one event ended and another began (Ben-Yakov and Henson, 2018). To determine the similarity between neural state boundaries and perceived event boundaries, we computed the boundary overlap defined as the number of timepoints where neural state and perceived event boundaries overlapped. We computed Spearman correlations (FDR corrected) to relate neural state duration and boundary overlap to age.

The median duration of neural states differed greatly between brain regions. In line with Geerligs et al (2022), we observed particularly short neural states in visual cortex, early auditory cortex, and somatosensory cortex. The longest states were observed in high-level regions such as the medial prefrontal gyrus and anterior portions of the lateral prefrontal cortex (Figure 1). There was a significant effect of age on neural state duration, with longer states with increasing age. This effect was strongest in the visual cortex and medial and lateral frontal cortex (Figure 2). Brain regions throughout the cortical hierarchy, showed significant overlap between neural state boundaries and perceived event boundaries. In particular, we observed that the anterior cingulate cortex, dorsal medial prefrontal cortex, frontal gyrus, and anterior insula show strong overlap. This suggests that neural state changes in these regions are most likely to underlie the experience of an event boundary. There was no significant effect of age on the overlap between neural state boundaries and perceived boundaries.

Our results show a hierarchy of neural states during movie viewing in a large lifespan cohort. We found that neural state durations were longer in older than younger adults in many cortical areas. This suggests that neural dedifferentiation with aging is not only reflected in decreased category specificity (Koen and Rugg, 2019) but also decreased temporal differentiation. Critically, the relationship between neural states and perceived event boundaries remained similar with age, suggesting that neural state boundaries that may underlie the experience of distinct events remains stable across the adult lifespan. This is in line with findings that show that older adults identify the same event boundaries as younger adults (Reagh et al., 2020).

Aging ¹

Activation (eg. BOLD task-fMRI) ²

Aging

FUNCTIONAL MRI

Segmentation

Other - naturalistic, movie, neural states