The hippocampal film-editor: sensitivity and specificity to event boundaries in continuous experience

Abstract

The function of the human hippocampus is normally investigated by experimental manipulation of discrete events. Less is unknown about what triggers hippocampal activity during more naturalistic, continuous experience. We hypothesized that the hippocampus would be sensitive to the occurrence of event boundaries, i.e. moments in time identified by observers as a transition between events. To address this, we analysed functional MRI data from two groups: one (N=253, 131 female) who viewed an 8.5min film and another (N=15, 6 female) who viewed a 120min film. We observed a strong hippocampal response at boundaries defined by independent observers, which was modulated by boundary strength (the number of observers that identified each boundary). In the longer film, there were sufficient boundaries to show that this modulation remained after covarying out a large number of perceptual factors. The hippocampus was the only brain region whose response showed a significant monotonic increase with boundary strength in both films, suggesting that modulation by boundary strength is selective to the hippocampus. This hypothesis-driven approach was complemented by a data-driven approach, in which we identified hippocampal-events as moments in time with the strongest hippocampal activity: The correspondence between these hippocampalevents and event boundaries was highly-significant, revealing that the hippocampal response is not only sensitive, but also specific to event boundaries. We conclude that the hippocampus plays an important role in segmenting the continuous experience that is typical of naturalistic settings.\n\nSignificance statementRecent years have seen the field of human neuroscience research transitioning from experiments with simple stimuli to the study of more complex and naturalistic experience. Nonetheless, our understanding of the function of many brain regions, such as the hippocampus, is based primarily on the study of brief, discrete events. As a result, we know little of what triggers hippocampal activity in real-life settings, when we are exposed to a continuous stream of information. When does the hippocampus \"decide\" to respond during the encoding of naturalistic experience? We reveal here that hippocampal activity measured by fMRI during film-watching is both sensitive and specific to event boundaries, identifying a potential mechanism whereby event boundaries shape experience by modulation of hippocampal activity.