August 2009
Volume 9, Issue 8
Free
Vision Sciences Society Annual Meeting Abstract  |   August 2009
An information theory approach to linking neuronal and behavioral temporal precision reveals sparse encoding and decoding underlying a rapid perceptual judgment
Author Affiliations
  • Ghose Geoffrey
    Neuroscience, University of Minnesota
  • Harrison Ian
    Neuroscience, University of Minnesota
Journal of Vision August 2009, Vol.9, 746. doi:https://doi.org/10.1167/9.8.746
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Ghose Geoffrey, Harrison Ian; An information theory approach to linking neuronal and behavioral temporal precision reveals sparse encoding and decoding underlying a rapid perceptual judgment. Journal of Vision 2009;9(8):746. https://doi.org/10.1167/9.8.746.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

A major challenge in understanding the neuronal mechanisms of cognition and perception is the possibility that a pattern or locus of neuronal activity may be correlated with a particular task, but not directly responsible for its execution. One approach is to simultaneously measure physiology and behavior in reaction time tasks. However, because neurons even in higher areas such as inferotemporal cortex have response latencies shorter than the fastest reaction times seen in perceptual tasks (∼150–200 ms), mean reaction time measures do not strongly constrain underlying physiological mechanisms. Moreover, reaction time distributions can be strongly affected by task or subject biases. Here we focus on measurements of behavioral performance that can place stronger constraints on physiological mechanisms: temporal precision and reliability. We quantify these measures by applying information theory to contingency tables between behavior and stimulus constructed at a variety of temporal resolutions and delays. The method integrates aspects of behavior that are traditionally considered separately: performance (% correct) and response timing. By applying this methodology to behavioral models based on the dynamics of a decision variable, including both diffusion and temporal probability summation, we find that the method's measures of dynamics are less susceptible to noise than traditional reaction time metrics. Finally, we demonstrate that the same methodology can also measure the precision of neuronal discharge. In particular, we analyzed the relationship of brief epochs of activity in area MT to both sensory events (encoding information rate) and behavioral choices (decoding information rate) during a rapid motion detection task. For this task, we find that the reliability of individual neurons over tens of milliseconds is nearly sufficient to explain behavioral performance. Thus, a consideration of temporal precision suggests cognitive tasks do not require extensive pooling of neuronal activity over either space or time.

Geoffrey, G. Ian, H. (2009). An information theory approach to linking neuronal and behavioral temporal precision reveals sparse encoding and decoding underlying a rapid perceptual judgment [Abstract]. Journal of Vision, 9(8):746, 746a, http://journalofvision.org/9/8/746/, doi:10.1167/9.8.746. [CrossRef]
Footnotes
 EY014989 NS5057091.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×