Purchase this article with an account.
Barry B. Lee; How ganglion cells code luminance and chromatic information in natural enviroments. Journal of Vision 2002;2(10):14. doi: 10.1167/2.10.14.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
There is growing interest in the way the visual system processes natural stimuli. We here examine the performance of primate retinal ganglion cells in response to such stimuli. Stimuli were confined to the temporal and chromatic domains and were derived from two contrasting environments, one typically northern European and the other a flower show. Cells' performance was evaluated in two ways; firstly, by investigating the variability of cell responses to repeated stimulus presentations, and secondly, by comparing the measured responses to the output of models developed for the cells. Both types of analysis yield a quantity called the coherence rate (in bit/s), which is related to information rate. Magnocellular (MC-) cells yielded coherence rates of 100 bit/s and more, while rates of parvocellular (PC-) cells were much lower. Ganglion cells driven by short-wavelength cones yielded coherence rates between those of PC- and MC-cells, From the (non-linear) cell models, we could show that for MC-cells information rate was almost exclusively related to luminance. PC-cells information was also dominated by achromatic content. This was due to the stimulus statistics; there is more luminance than chromaticity variation in natural environments. Even at the flower show, only about one sixth of the PC-cell's coherence rate derived from the chromatic content. Chromatic information appeared dominated by frequencies below 10 Hz. For S-cone cells, information was also restricted to lower frquencies (<20 Hz). For frequencies below 2–3 Hz, PC-cell signals contained considerably more power than those of MC-cells. These results allow us to quantify how information in luminance and spectral aspects of the stimuli is distributed amongst the different classes of ganglion cells and lead toward development of full spatiotemporal models as preprocessing modules for studies of higher visual processing.
This PDF is available to Subscribers Only