Fine spatial information represented in a population of retinal ganglion cells

Frederick S. Soo, Gregory W. Schwartz, Kolia Sadeghi, Michael J. Berry

Research output: Contribution to journalArticlepeer-review

19 Scopus citations

Abstract

Detailed measurement of ganglion cell receptive fields often reveals significant deviations from a smooth, Gaussian profile. We studied the effect of these irregularities on the representation of fine spatial information in the retina. We recorded from nearby clusters of ganglion cells, testing their ability to determine the location of small flashed spots, and we compared the results to the prediction of a Gaussian receptive field model derived from reverse correlation. Despite considerable receptive field overlap, almost all ganglion cell pairs signaled nearly independently. For groups of five cells with highly overlapping receptive fields, the measured light-evoked currents encoded ∼33% more information than predicted by the Gaussian receptive field model. Including measured local irregularities in the receptive field model increased performance to the level observed experimentally. These results suggest that instead of being an unavoidable defect, irregularities may be a positive design feature of population neural codes. Copyright

Original languageEnglish (US)
Pages (from-to)2145-2155
Number of pages11
JournalJournal of Neuroscience
Volume31
Issue number6
DOIs
StatePublished - Feb 9 2011

All Science Journal Classification (ASJC) codes

  • General Neuroscience

Fingerprint

Dive into the research topics of 'Fine spatial information represented in a population of retinal ganglion cells'. Together they form a unique fingerprint.

Cite this