MIT CogNet, The Brain Sciences ConnectionFrom the MIT Press, Link to Online Catalog
SPARC Communities
Subscriber : Stanford University Libraries » LOG IN

space

Powered By Google 
Advanced Search

 

Neural Correlates of Event Segmentation

 Jeffrey M. Zacks, Todd S. Braver, Margaret A. Sheridan, David I. Donaldson, Abraham Z. Snyder, John M. Ollinger, Randy L. Buckner and Marcus E. Raichle
  
 

Abstract:
Temporal structure plays a major role in human understanding of everyday events. It has been shown that, when asked, observers are able to segment ongoing activity into temporal parts that are reliable, meaningful, and correlated with ecologically relevant features of the action. Here we present evidence that a network of brain regions is tuned to the perceptual part structure of ongoing activity. Participant watched movies of everyday activities (e.g., making a bed) while local brain activity was recorded with functional MRI (fMRI). Following these passive viewing scans, participants watched the same movies and were asked to tap a button to mark the boundaries between natural and meaningful units. Boundaries were identified at both a coarse and fine temporal grain (during separate scans), and these boundaries were used to define "trials" in an event-related fMRI analysis. A network of brain regions showed reliable transient changes in activity, during the passive viewing runs, at the location of event segment boundaries. The most prominent regions of activity were a bilateral parieto-occipital region (including human MT complex) and a right superior frontal region (human frontal eye field). Areas that were active during passive viewing were also active during the active segmentation runs. In most regions, activity was greater for coarse-unit boundaries than large-unit boundaries. These results strongly indicate that a network of cortical regions tracks observers' temporal segmentation of ongoing activity. Moreover, this appears to be an naturally concomitant component of ongoing perception, not tied to a particular task.

 
 


© 2010 The MIT Press
MIT Logo