Responses of primate retinal ganglion cells to perimetric stimuli

Abstract

Purpose: Perimetry is used clinically to assess glaucomatous ganglion cell loss. It has been proposed that frequency-doubling stimuli are better than the conventional size III perimetric stimulus in preferentially stimulating magnocellular (M) versus parvocellular (P) ganglion cells. However, little is known about how primate ganglion cells respond to perimetric stimuli. The authors recorded contrast responses of M and P ganglion cells to size III and frequency-doubling stimuli and compared contrast gain of M and P cells to these stimuli to assess the ability of these stimuli to preferentially stimulate M versus P cells.

Methods: Data were recorded from 69 macaque retinal ganglion cells, by an in vivo preparation, at eccentricities of 5° to 15°. The size III stimulus was a circular luminance increment 26 min arc in diameter, 200 ms in duration. The frequency-doubling stimulus was a sinusoidal grating (0.5 cyc/deg) temporally modulated in counterphase at 13 Hz. A Michaelis-Menten function was fit to each cell's contrast responses to assess contrast gain.

Results: For both size III and frequency-doubling stimuli, ganglion cell responses increased linearly at low contrasts, and then the increase slowed at high contrasts (saturation). The mean (± SE) difference in estimated log contrast gain between M and P cells for the size III stimulus was significantly higher than that for the frequency-doubling stimulus (1.24 ± 0.09 vs. 0.89 ± 0.13; P < 0.01).

Conclusions: The size III stimulus was superior to the frequency-doubling stimulus in preferentially stimulating M cells versus P cells.

Figure 1.

Response histograms for the size III stimulus, from four representative cells. Each histogram has 10-ms bin width and is averaged across 30 trials. Bottom: time course of the 200-ms stimulus presentation. Right: stimulus contrasts.

Figure 2.

Response histograms for the frequency-doubling stimulus from the same cells shown in Figure 1. Each histogram has 2.4-ms bin width and is averaged across 100 trials. Bottom: time course of the 13-Hz stimulus presentation over two cycles. Right: stimulus contrasts.

Figure 3.

Contrast responses for the size III stimulus (top) and the frequency-doubling stimulus (bottom) for the same cells illustrated in Figures 1 and 2. Circles: spike rates for different contrasts. Curves: Michaelis-Menten functions fit to the data. For consistency between stimuli, data are shown only up to 100% stimulus contrast. Right: size III and frequency-doubling stimuli.

Figure 4.

Left: mean contrast gain for size III and frequency-doubling stimuli, for M and P cells. SD (error bars) was similar for all four types of contrast gain. Right: mean (± SE) differences in log contrast gain for M cells minus P cells, for size III and frequency-doubling stimuli. The difference was significantly larger for the size III stimulus than for the frequency-doubling stimulus, and the size III stimulus was superior to the frequency-doubling stimulus in preferentially stimulating M cells.

Figure 5.

Scatter plot of contrast gains for the 25 cells tested with both stimuli (small solid symbols) and the means and 95% confidence limits for all 69 cells (large open symbols and error bars). The diagonal line has a slope of 1 and passes through the mean for the M cells; if the frequency-doubling stimulus had been superior to the size III stimulus in stimulating M versus P cells, the contrast gains for P cells would have fallen above this line.

Figure 6.

Ganglion cell responses to the size III stimulus when the stimulus was offset from the receptive field center. Symbols: mean spike rate for each stimulus contrast and offset.