Learning Naturalistic Temporal Structure in the Posterior Medial Network

Abstract

The posterior medial network is at the apex of a temporal integration hierarchy in the brain, integrating information over many seconds of viewing intact, but not scrambled, movies. This has been interpreted as an effect of temporal structure. Such structure in movies depends on pre-existing event schemas, but temporal structure can also arise de novo from learning. Here we examined the relative role of schema-consistent temporal structure and arbitrary but consistent temporal structure on the human posterior medial network. We tested whether, with repeated viewing, the network becomes engaged by scrambled movies with temporal structure. Replicating prior studies, posterior medial regions were immediately locked to stimulus structure upon exposure to intact but not scrambled movies. However, for temporally structured scrambled movies, functional coupling within the network increased across stimulus repetitions, rising to the level of intact movies. Thus, temporal structure is a key determinant of network dynamics and function in the posterior medial network.

Temporal integration of narrative information in a hippocampal amnesic patient

“…Even if their online processing of stimuli is preserved, hippocampal amnesics are nonetheless unable to retrieve episodic information at a later time, either due to disrupted encoding or retrieval. Studies of functional connectivity between the hippocampus/MTL and default network during naturalistic input have emphasized the interaction between these structures in service of memory encoding and retrieval in the healthy brain (van Kesteren et al, 2010;Chen et al, 2016;Bonasia et al, 2018;Aly et al, 2018), and hippocampal responses at the boundaries between events are theorized to support consolidation of event information (Ben-Yakov et al, 2013;Baldassano et al, 2017;Ben-Yakov and Henson, 2018). Interestingly, MTL damage does not impact connectivity among default network nodes during resting state, but does alter connectivity between the MTL and default network areas (Hayes et al, 2012); in fact, it has been proposed that abnormal hippocampal interactions with other brain regions are the cause of behavioral deficits in amnesia, as opposed to hippocampal damage per se (Argyropoulos et al, 2019).…”

Measuring shared responses across subjects using intersubject correlation

“…If subject A receives the same stimulus multiple times, the correlation between these multiple instances will be sensitive to idA(t), to the extent that the subject-specific processes are stable over multiple exposures. Experience, however, changes how we process stimuli (e.g., Lahav et al, 2007;Engel et al, 2012;Aly et al, 2018). Before measuring intrasubject correlation, one should consider two caveats: first, being exposed repeatedly to the same stimulus leads to habituation (Grill-Spector et al, 2006); second, some idiosyncratic processes are unreliable in their timing, and would thus still fail to register as intrasubject correlation in repeated sessions.…”

Episodic memory: mental time travel or a quantum `memory wave' function?

“…For example, prior experiences with shared or overlapping contextual properties (e.g., other experiences that occurred in similar spatial or social settings, shared similar goals, etc.) could be leveraged to form schemas or situation models that help guide behaviors and expectations according to the current perceived context (Ranganath and Ritchey, 2012;Baldassano et al, 2018;Aly et al, 2018).…”

“…Essentially, slowly drifting thoughts may be used to "index" more rapidly drifting thoughts that are temporally proximal. This notion, that our brain maintains parallel mental representations drifting at different time scales, has been well-characterized in a series of elegant fMRI and ECoG studies by Uri Hasson and colleagues (Hasson et al, 2008;Lerner et al, 2011;Honey et al, 2012;Aly et al, 2018). Subsequent work by the same group has shown that this hierarchy of drifting mental representations plays a central role in how we remember continuous experiences and segment our experiences into discrete events (Baldassano et al, 2017).…”