April 2009
Volume 50, Issue 13
Free
ARVO Annual Meeting Abstract  |   April 2009
Adaptation of Retinal Ganglion Cells to Natural-image Contrast
Author Affiliations & Notes
  • X. Cao
    Biomedical Engineering, University of Southern California, Los Angeles, California
  • D. K. Merwine
    Division of Biological and Health Science, University of Pittsburgh at Bradford, Bradford, Pennsylvania
  • N. M. Grzywacz
    Biomedical Engineering, University of Southern California, Los Angeles, California
  • Footnotes
    Commercial Relationships  X. Cao, None; D.K. Merwine, None; N.M. Grzywacz, None.
  • Footnotes
    Support  This work was funded by National Eye Institute grants EY11170 and EY016093 to NMG
Investigative Ophthalmology & Visual Science April 2009, Vol.50, 1415. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      X. Cao, D. K. Merwine, N. M. Grzywacz; Adaptation of Retinal Ganglion Cells to Natural-image Contrast. Invest. Ophthalmol. Vis. Sci. 2009;50(13):1415.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose: : Retinal ganglion cells (RGCs) are known to adapt to the statistics of spatial and temporal contrasts when using artificial images. Here, we investigate whether RGCs also adapt to these statistics in the natural environment.

Methods: : We used natural-image sets with different mean RMS contrasts to stimulate rabbit RGCs (one image every 30 ms). A mixture-of- exponentials model is used to estimate the time constants.

Results: : 1. RGCs show contrast adaptation to natural stimuli. Three different time constants were found in the switch from low to high contrasts. The fastest time constant (TC) was < 0.2 s, termed contrast gain control. A second process was fast adaptation (TC: 0.7 ~ 4.5 s with a mode of 1.3 s). A third process showed slow adaptation (TC: 10 ~ 45 s with a mode of 20 s). In contrast, adaptation from high to low contrasts exhibited only one time constant (10 ~ 40 s with a mode of 15 s).2. Across a population of 125 cells, 82 (66%) yielded adaptation to the switch from low to high contrasts. Of them, 27 (33%) showed slow adaptation, 24 (29%) showed fast adaptation, and 31 (38%) showed both fast and slow adaptation. In addition, only 19 cells out of 82 (23%) displayed adaptation from high to low contrasts.3. The percentage of transient RGCs showing contrast adaptation (69 from 96 cells - 72%) was statistically significantly higher (chi^2 test, p < 0.01) than that of sustained RGCs (9 from 23 cells - 39%). In turn, On cells relatively rarely showed fast adaptation (10 from 38 cells - 26%), compared to Off cells (27 from 50 cells - 54%).4. Two kinds of behavior were observed when we varied the contrast difference between the low- and high-contrast sets of natural images. The first group (63% of cells) showed adaptation to almost all contrast changes. The second group (37% of cells) yielded adaptation only to large contrast differences.5. As we increased the difference between the low- and high-contrast sets, the slow time constants decreased, and the initial and steady-state adaptive responses increased, but their ratio remained statistically constant.

Conclusions: : A large fraction but not all rabbit RGCs show contrast adaptation to natural images. This adaptation can occur with three different time constants, indicating multiple mechanisms. Different cell types use a distinct subset of these mechanisms, suggesting that different visual tasks have distinct requirements of RGC contrast adaptation. Finally, that the ratio between initial and steady-state responses remains constant imposes strong constraints on adaptation mechanisms. Simple linear gain-decay models cannot account for such constant ratio.

Keywords: adaptation: pattern • electrophysiology: non-clinical • ganglion cells 
×
×

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.

×