Quantitative aspects of sensitivity and summation in the cat retina

Abstract

SUYMMRY1. Properties of the central response mechanism of on-centre ganglion cells in the cat retina were studied by recording, from optic tract fibres, responses evoked by stimuli modulated with time in a sinusoidal or squarewave fashion.2. The shape of averaged square-wave responses resulting from the central mechanism alone was identified. This shape was identical from one cell to another. Such an identification permits the early recognition of peripheral antagonism.3. Threshold sensitivity for a sinusoidal stimulus was determined for fifty cells along one horizontal and vertical axis, passing through the most sensitive portion of the receptive field. These sensitivity profiles were described in terms of a central segment of constant maximum sensitivity (uniform centre) and sloping outer segments of exponentially decreasing sensitivity (exponential annulus). The dimensions of the uniform centre (horizontal axis x vertical axis) varied from 0.10 x 0.10 to 2.5 x 2-20, the half width of the exponential annulus ranged from 0.10 to 0.630. 18 B. G. CLELANTD AND CHRISTINA ENROTH-CUGELL central mechanism alone contributed to the response. These area-sensitivity curves did not show any decrease in sensitivity at larger diameters. 6. The shape of the area-sensitivity curve, and hence the extent of the -ummating area, was found to be independent of the state of adaptation.7. For any one cell the shapes of the area-adaptation and area-sensitivity curves were shown to be identical, indicating that adapting flux and stimulus flux are independent of distribution over the same defined limits.8. The sensitivity of combinations of small disconnected areas of the receptive field was found to be equal to the sum of their individual sensitivities.

Apparent movement and metacontrast: A note on Kahneman’s formulation

“…Thus, a lateral inhibition model would be in no way inconsistent with the metacontrast effects obtained in this study. There is some evidence that with increase in energy, inhibition from the periphery of a receptive field is "turned on" and begins to influence the center of a receptive field in the absence of direct stimulation to the flanks (Cleland & Enroth-Cugell, 1968); increases in inhibition of this type might playa role in the general increase (with the exception of Experiment 3) in "sharpness" of the metacontrast functions with increasing energy. In terms of maxima, there were no consistent trends towards or away from the origin with increases in energy, as might be predicted from rate constant and amplitude changes in excitation and inhibition functions; however, as noted, these changes would predict a tangled picture, with a number of effects possibly cancelling each other.…”

“…Second, there ought to be fairly complicated changes in the metacontrast function with changes in the energy level of targets and masks, which are themselves of relatively equal energy, since there seem to be fairly complicated changes in excitation and inhibition with changes in the energy levels of stimuli. With increase in energy, a number of things may be expected to occur: excitation might be expected to decrease in gain and to develop a faster rate of response (Sperling & Sondhi, 1968;Matin, 1968;Fuortes & Hodgkin, 1964); inhibition might be expected also to decrease in gain and to increase in rate, although not necessarily in parallel with the changes in excitation; peripheral antagonism might be expected to contribute more and more to the center of the receptive field when the target is presented alone (Cleland & Enroth-Cugell, 1968). At the same time, if there were not sensitivity and rate constant changes with increase in energy, the maxima of the metacontrast functions might be expected to occur at larger and larger values of~t, due to the development of fuller inhibitory and excitatory responses, while the amplitude of the metacontrast function might decrease, due to the differentially higher level of excitation minus inhibition at maximum masking.…”

Informational and neural adaptation curves are asynchronous

“…As I mentioned earlier, the time course of adaptation to constant-intensity stimuli (small spots of light) of retinal ganglion cells is somewhat slower than that of spiral ganglion cells to sounds. Firing in at least some central retinal (foveal) ganglion cells declines to background rate over a time period of at least 0.5 sec (Cleland & Enroth-Cugell, 1968). Table 3 and Figure 3 present the results of fitting Equation 4 to data from a single, representative cat retinal ganglion cell of this type read from a graph presented by Cleland and Enroth-Cugell.…”

“…Estimating Fmax to be about 300 and Fmin to be about 40 for the auditory data in Figure 1 and their analysis by Yates et al (1985), we obtain a channel capacity of roughly (300-40)/127.8 = 2.03 bits for sound intensity, very close to the value of 2.1 reported by Garner (1962) and Miller (1956) and assumed here for the calculation of stimulus equivocation. Similarly, for the retinal ganglion cell reported by Cleland and Enroth-Cugell (1968), Fmax = about 274, Fmin = about 0, and k = 210, so that channel capacity is predicted to be about (274-0)/210 = 1.3. This is somewhat smaller than the maximum value of roughly 2.0 bits usually quoted for light intensity (see, e.g., Garner, 1962); but it is close to the maximum measured in the present experiment (1.54 bits), and, given …”

“…For example, Lockhead (1966) varied both stimulus duration (200 vs. 8 msec) and luminance (no filter vs. 1 log unit neutral density filter) in absolute identification of line This research was supported by a grant from the Natural Sciences and Engineering Research Council of Canada. I am grateful to Ken Norwich for many discussions regarding this work and for his insightful comments on this paper; for bringing the papers by Yates, Robertson, and Johnstone (1985) and Cleland and Enroth-Cugell (1968) to my attention; for supplying the simplex optimizationalgorithm; and for providing access to his own unpublished work. I thank Odie Geiger, Steven Lindsay, and Kelly Davidson for help with running subjects, and Shuji Mori for help with the setup for the visual stimuli.…”