Purchase this article with an account.
H Sun, BB Lee, L Rüttiger; Grating Vernier thresholds as a Function of Temporal and Spatial Frequency: A Unitary Retinal Mechanism? . Invest. Ophthalmol. Vis. Sci. 2002;43(13):2836.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: Levi (Cur. Biol., 1996, vol. 6, pp1020) showed that with grating targets of different spatial frequencies, vernier thresholds as a function of temporal frequency superimposed when expressed in terms of grating phase. Is this feature present at the retinal level, or is it derived from cortical processing? To test between these alternatives, we estimated spatial accuracy of ganglion cell responses to drifting gratings of various spatial, temporal frequencies and contrasts, and compared the ganglion cell behavior to psychophysical vernier threshold. Methods: The stimuli were drifting gratings of various spatial (0.1-3.2 cpd) and temporal (0.5-26 Hz) frequencies and contrasts (2.5-80%). In physiological experiments, parafoveal ganglion cell responses to gratings were obtained from the retinae of anesthetized macaques. The spatial variability of cell responses was assessed from response phase using fourier analysis and other neurometric techniques. In psychophysical experiments, vernier thresholds in parafovea were estimated with a staircase method. Results: Positional accuracy of MC-cell responses as a function of temporal frequency superimposed for different spatial frequencies (0.1-1.6cpd) when expressed as grating phase. The results showed a similar pattern to our psychophysical vernier thresholds in parafovea, and to Levi's foveal data. Transient MC-cells could provide accurate positional information even for slowly moving gratings. The positional information delivered by PC-cells was less accurate, and deteriorated rapidly as temporal frequency increased. As grating contrast increased, positional reliability of MC-cell responses increased with a slope of ∼1.0. Psychophysical vernier performance improved with a shallower slope of ∼0.5 and showed saturation at 40-80% contrast. PC-cells showed weak responses to gratings of lower than 40% contrast. Conclusion: These results suggest that the relationship of vernier thresholds to spatial and temporal frequency is already inherent in the retinal signal. Our data are consistent with MC-cell responses provide important information to vernier performance with drifting gratings, but differences in detail (e.g., contrast slope) imply additional cortical processing.
This PDF is available to Subscribers Only