Transfer characteristics of excitation and inhibition in cat retinal ganglion cells

Illusory Shifts in Spatial Frequency Caused by Temporal Modulation

It is known that temporal modulation of a sinusoidal grating of low spatial frequency causes an increase in the apparent spatial frequency of the grating. A possible explanation for the illusion is proposed. Temporal modulation would favour channels which respond only transiently to prolonged presentation of a grating. These channels may be the human analogues of Y-cells found in the cat retina. Y-cells respond nonlinearly to gratings, and the nonlinearity may be the root of the apparent spatial-frequency illusions.

Stimulus size dependence of information transfer from retina to thalamus

Relay cells in the mammalian lateral geniculate nucleus (LGN) are driven primarily by single retinal ganglion cells (RGCs). However, an LGN cell responds typically to less than half of the spikes it receives from the RGC that drives it, and without retinal drive the LGN is silent (Kaplan and Shapley, 1984). Recent studies, which used stimuli restricted to the receptive field (RF) center, show that despite the great loss of spikes, more than half of the information carried by the RGC discharge is typically preserved in the LGN discharge (Sincich et al., 2009), suggesting that the retinal spikes that are deleted by the LGN carry less information than those that are transmitted to the cortex. To determine how LGN relay neurons decide which retinal spikes to respond to, we recorded extracellularly from the cat LGN relay cell spikes together with the slow synaptic ('S') potentials that signal the firing of retinal spikes. We investigated the influence of the inhibitory surround of the LGN RF by stimulating the eyes with spots of various sizes, the largest of which covered the center and surround of the LGN relay cell's RF. We found that for stimuli that activated mostly the RF center, each LGN spike delivered more information than the retinal spike, but this difference was reduced as stimulus size increased to cover the RF surround. To evaluate the optimality of the LGN editing of retinal spikes, we created artificial spike trains from the retinal ones by various deletion schemes. We found that single LGN cells transmitted less information than an optimal detector could.

Central factors contributing to para-contrast modulation of contour and brightness perception

Following up on a prior study of contour and brightness processing in visual masking (Breitmeyer et al., 2006), we investigated the effects of a binocular and dichoptic para-contrast masking on the visibility of the contour and brightness of a target presented to the other eye. Combined, the results support the contributions of several cortical processes to para-contrast: (1) two central sources of inhibition, one long-latency and prolonged and the other short-latency and brief; (2) binocular rivalry suppression; and (3) a facilitatory effect peaking at different SOAs for the contour and the brightness tasks, reflecting; (4) known properties of two separate cortical systems, one a fast contour-processing pathway and the other a slower brightness-processing pathway.

Meta- and paracontrast reveal differences between contour- and brightness-processing mechanisms

We investigated meta- and paracontrast masking using tasks requiring observers to judge the surface brightness or else the contours of target stimuli. The contour task revealed strongest metacontrast at SOAs shorter than those obtained for the brightness task. Paracontrast revealed related temporal differences between the tasks. Additionally, the paracontrast results support the existence not only of prolonged inhibitory effects but also of facilitatory effects. The combined results comport with the existence of cortical mechanisms for: (i) fast contour processing, (ii) slow surface-brightness processing, (iii) prolonged inhibition, and (iv) facilitation.

The receptive fields of cat retinal ganglion cells in physiological and pathological states: where we are after half a century of research

Studies on the receptive field properties of cat retinal ganglion cells over the past half-century are reviewed within the context of the role played by the receptive field in visual information processing. Emphasis is placed on the work conducted within the past 20 years, but a summary of key contributions from the 1950s to 1970s is provided. We have sought to review aspects of the ganglion cell receptive field that have not been featured prominently in previous review articles. Our review of the receptive field properties of X- and Y-cells focuses on quantitative studies and includes consideration of the function of the receptive field in visual signal processing. We discuss the non-classical as well as the classical receptive field. Attention is also given to the receptive field properties of the less well-studied cat ganglion cells-the W-cells-and the effect of pathology on cat ganglion cell properties. Although work from our laboratories is highlighted, we hope that we have given a reasonably balanced view of the current state of the field.

100 years of Benham's top in colour science

For 100 years Benham's top has been a popular device demonstrating pattern-induced flicker colours (PIFCs). Results of early and recent investigations on PIFCs are reported and show that the phenomenon originates in phase-sensitive lateral interactions of modulated neural activity in the retina followed by additional spatial interactions in the visual cortex behind the locus of binocular fusion. Colour matches with normal colour stimuli indicate that S/(M + L) opponent neurons are involved. Dichromats do not find matching stimuli for all PIFCs. PIFCs may become useful in medical diagnosis. The phenomenon is interpreted as a side effect of a neural mechanism providing colour constancy under normal stimulus conditions.

The effects of temporal modulation and spatial location on the perceived spatial frequency of visual patterns

The perceived spatial frequency of a visual pattern can increase when a pattern drifts or is presented at a peripheral visual field location, as compared with a foveally viewed, stationary pattern. We confirmed previously reported effects of motion on foveally viewed patterns and of location on stationary patterns and extended this analysis to the effect of motion on peripherally viewed patterns and the effect of location on drifting patterns. Most central to our investigation was the combined effect of temporal modulation and spatial location on perceived spatial frequency. The group data, as well as the individual sets of data for most observers, are consistent with the mathematical concept of separability for the effects of temporal modulation and spatial location on perceived spatial frequency. Two qualitative psychophysical models suggest explanations for the effects. Both models assume that the receptive-field sizes of a set of underlying psychophysical mechanisms monotonically change as a function of temporal modulation or visual field location, whereas the perceptual labels attached to a set of channels remain invariant. These models predict that drifting or peripheral viewing of a pattern will cause a shift in the perceived spatial frequency of the pattern to a higher apparent spatial frequency.

Sensitivity of macaque retinal ganglion cells to chromatic and luminance flicker

1. We have studied the sensitivity of macaque retinal ganglion cells to sinusoidal flicker. Contrast thresholds were compared for stimuli which alternated only in luminance ('luminance flicker') or chromaticity ('chromatic flicker'), or which modulated only the middle- or long-wavelength-sensitive cones ('silent substitution'). 2. For luminance flicker, the lowest thresholds were those of phasic, non-opponent ganglion cells. Sensitivity was maximal near 10 Hz. 3. Tonic, cone-opponent ganglion cells were relatively insensitive to luminance flicker, especially at low temporal frequencies, but were sensitive to chromatic flicker, thresholds changing little from 1 to 20 Hz. Those with antagonistic input from middle- and long-wavelength-sensitive (M- and L-) cones had a low threshold to chromatic flicker between red and green lights. Those with input from short-wavelength-sensitive (S-) cones had a low threshold to chromatic flicker between blue and green. Expressed in terms of cone contrast, the S-cone inputs to blue on-centre cells had higher thresholds than M- and L-cone inputs to other cell types. 4. Phasic, non-opponent cells responded to high-contrast red-green chromatic flicker at twice the flicker frequency. This frequency-doubled response is due to a non-linearity of summation of M- and L-cone mechanisms. It was only apparent at cone contrasts which were above threshold for most tonic cells. 5. M- or L-cones were stimulated selectively using silent substitution. Thresholds of M- and L-cone inputs to both red and green on-centre cells were similar. This implies that these cells' sensitivity to chromatic flicker is derived in equal measure from centre and surround. Thresholds of the isolated cone inputs could be used to predict sensitivity to chromatic flicker. The high threshold of these cells to achromatic contrast is thus, at least in part, due to mutual cancellation by opponent inputs rather than intrinsically low sensitivity. 6. Thresholds of M- and L-cone inputs to phasic cells were similar at 10 Hz, and were comparable to those of tonic cells, suggesting that at 1400 td cone inputs to both cell groups are of similar strength. 7. The modulation transfer function of phasic cells to luminance flicker was similar to the detection sensitivity curve of human observers who viewed the same stimulus. For chromatic flicker, at low temporal frequencies thresholds of tonic cells (red or green on-centre cells in the case of red-green flicker or blue on-centre cells in the case of blue-green flicker) approached that of human observers. We propose the different cell types are the substrate of different channels which have been postulated on the basis of psychophysical experiments. 8. At frequencies of chromatic flicker above 2 Hz, human sensitivity falls off steeply whereas tonic cell sensitivity remained the same or increased. This implies that high-frequency signals in the chromatic, tonic cell pathway are not available to the central pathway respons

The physiological basis of heterochromatic flicker photometry demonstrated in the ganglion cells of the macaque retina

1. Heterochromatic flicker photometry is a way of measuring the spectral sensitivity of the human eye. Two lights of different colour are sinusoidally alternated at, typically, 10-20 Hz, and their relative intensities adjusted by the observer until the sensation of flicker is minimized. This technique has been used to define the human photopic luminosity, or V lambda, function on which photometry is based. 2. We have studied the responses of macaque retinal ganglion cells using this stimulus paradigm. The responses of the phasic ganglion cells go through a minimum at relative radiances very similar to that predicted from the V lambda function. At this point, defined as equal luminance, an abrupt change in response phase was observed. A small residual response at twice the flicker frequency was apparent under some conditions. 3. The spectral sensitivity of parafoveal phasic cells measured in this way corresponded very closely to that of human observers minimizing flicker on the same apparatus. 4. Minima in phasic cell activity were independent of flicker frequency, as is the case in the psychophysical task. 5. The response minima of phasic cells obey the laws of additivity and transitivity which are important characteristics of heterochromatic flicker photometry. 6. As the relative intensities of the lights were altered responses of tonic, spectrally opponent cells usually underwent a gradual phase change with vigorous responses at equal luminance. The responses of tonic cells treated individually or as a population could not be related to the V lambda function in any meaningful way. 7. We conclude that the phasic, magnocellular cell system of the primate visual pathway underlies performance in the psychophysical task of heterochromatic flicker photometry. It is likely that other tasks in which spectral sensitivity conforms to the V lambda function also rely on this cell system.

The dynamics of the cat retinal X cell centre

1. The dynamics of the centre mechanism of individual cat X retinal ganglion cells is investigated. The visual stimuli consist of temporal contrast modulation of stationary patterns. In order to study the response of the centre mechanism, patterns were either sine gratings of high spatial frequency or small circular spots positioned over the receptive-field centre. 2. Responses to contrast reversal are approximately linear. However, as the modulation depth of the stimulus increases, responses become more transient. Ganglion cell responses show this phenomenon at moderate contrasts (e.g. 0.1), which do not elicit discharges that approach the maximum firing rate of the ganglion cell. 3. A sequence of dynamical models are constructed from responses elicited by sum-of-sinusoids modulation of the spatial pattern. The first model is strictly linear. It consists of a series of low-pass filters and a single high-pass filter. The linear model predicts the approximate shape of the step response, but does not account for the change in shape of the response as a function of modulation depth. 4. The second model, a quasi-linear model, allows the 'linear' dynamics to vary slowly with a neural measure of contrast. The main effect of high contrast is a shorter time constant in the high-pass filter. This model accounts qualitatively for the increased transience of the response, but fails to predict the magnitude of the effect at higher modulation depths. 5. In the third model, the transfer characteristics of the centre response adjust rapidly as contrast changes. This intrinsically non-linear model provides excellent agreement with observed response to steps and more complex modulation patterns. 6. The non-linearity necessitated by a voltage-to-spikes transduction is analysed quantitatively. In most ganglion cells, a simple truncation at 0 impulses/s (and no saturation) explains the changes in apparent gain and mean firing rate that occur as modulation depth is increased. A non-linear voltage-to-spike transduction per se cannot account for the observed effect of contrast on dynamics. 7. The parameters of the dynamical model are measured for a population of twenty-seven X ganglion cells (nineteen on-centre and eight off-centre). The low-pass stage and the strength of the high-pass stage are relatively uniform across the population. The over-all gain and the dynamics of the high-pass stage vary substantially across the population, but show no consistent dependence on the on-off distinction or on retinal location. Some implications of this variability for retinal function are discussed.

Spatiotemporal frequency responses of cat retinal ganglion cells

Spatiotemporal frequency responses were measured at different levels of light adaptation for cat X and Y retinal ganglion cells. Stationary sinusoidal luminance gratings whose contrast was modulated sinusoidally in time or drifting gratings were used as stimuli. Under photopic illumination, when the spatial frequency was held constant at or above its optimum value, an X cell's responsivity was essentially constant as the temporal frequency was changed from 1.5 to 30 Hz. At lower temporal frequencies, responsivity rolled off gradually, and at higher ones it rolled off rapidly. In contrast, when the spatial frequency was held constant at a low value, an X cell's responsivity increased continuously with temporal frequency from a very low value at 0.1 Hz to substantial values at temporal frequencies higher than 30 Hz, from which responsivity rolled off again. Thus, 0 cycles X deg-1 became the optimal spatial frequency above 30 Hz. For Y cells under photopic illumination, the spatiotemporal interaction was even more complex. When the spatial frequency was held constant at or above its optimal value, the temporal frequency range over which responsivity was constant was shorter than that of X cells. At lower spatial frequencies, this range was not appreciably different. As for X cells, 0 cycles X deg-1 was the optimal spatial frequency above 30 Hz. Temporal resolution (defined as the high temporal frequency at which responsivity had fallen to 10 impulses X s-1) for a uniform field was approximately 95 Hz for X cells and approximately 120 Hz for Y cells under photopic illumination. Temporal resolution was lower at lower adaptation levels. The results were interpreted in terms of a Gaussian center-surround model. For X cells, the surround and center strengths were nearly equal at low and moderate temporal frequencies, but the surround strength exceeded the center strength above 30 Hz. Thus, the response to a spatially uniform stimulus at high temporal frequencies was dominated by the surround. In addition, at temporal frequencies above 30 Hz, the center radius increased.

Identification and estimation algorithm for stochastic neural system

An algorithm for the estimation of stochastic processes in a neural system is presented. This process is defined here as the continuous stochastic process reflecting the dynamics of the neural system which has some inputs and generates output spike trains. The algorithm proposed here is to identify the system parameters and then estimate the stochastic process called neural system process here. These procedures carried out on the basis of the output spike trains which are supposed to be the data observed in the randomly missing way by the threshold time function in the neural system. The algorithm is constructed with the well-known Kalman filters and realizes the estimation of the neural system process by cooperating with the algorithm for the parameter estimation of the threshold time function presented previously (Nakao et al., 1983). The performance of the algorithm is examined by applying it to the various spike trains simulated by some artificial models and also to the neural spike trains recorded in cat's optic tract fibers. The results in these applications are thought to prove the effectiveness of the algorithm proposed here to some extent. Such attempts, we think, will serve to improve the characterizing and modelling techniques of the stochastic neural systems.

Parameter estimation of the threshold time function in the neural system

An algorithm for parameter estimation is presented for the neural system model. Because of its firing mechanism analogous to that of the model based on the first time crossing problem, this problem is solved numerically for our model according to the results of Kostyukov et al. (1981). We propose the algorithm that estimates the parameters of the model considering the equivalence between the probability density function of the 1st crossing time and that of the interspike interval, which is derived from the interspike interval histogram by making use of the spline function technique. The ability of the algorithm is ensured by the application to the simulated interspike interval data. The parameter estimation is carried out also for the practical neural data recorded in the cat's optic tract fibers in both the spontaneous and the stimulated cases. These applications will show the effectiveness of the algorithm in practical cases.

Visual masking: a unified approach

The suppression effect observed in masking is assumed to be due to neural inhibition between two stimuli that interact spatially or within a narrow span of time in the fashion demonstrated in simultaneous brightness-contrast experiments. Three principles are brought together in a comprehensive mathematical model of visual masking. The three principles are: lateral inhibition, the integrated visual response function (VRF), and stimulus decay after its offset over a limited period of time (up to about 100 ms). Block's law is extended to apply to the integration of the VRF, including decay. The three principles combined are sufficient to explain well-known masking phenomena such as metacontrast, and forward and backward masking. The mathematical model presented demonstrates clearly the common underlying basis of all masking. The validity of the lesser documented decay after stimulus offset, a necessary assumption to explain masking effects with stimuli that are delayed relative to each other, is demonstrated with experimental evidence.

Reconstruction of cone-system contributions to responses of colour-opponent neurones in monkey lateral geniculate

Responses of colour-opponent X-cells to intensity-modulation at various wavelengths were obtained in the lateral geniculate nucleus (LGN) of the anaesthetized (N2O/O2) rhesus monkey. The gaussian white noise (GWN) analysis method was used to describe the stimulus-response relationship. Two different methods were used to estimate sign and relative strength of the response contribution of each of the three known cone systems as a function of time. Both methods revealed that, in contrast to the well-known variability in gain and sign, the time course of the cone-type contributions was remarkably stereotyped in all cells. Surround-mediated cone-type contributions appeared to have a consistently longer delay than centre-mediated inputs. Response contributions from different types of cone appeared to add linearly in LGN neurones. Apart from rectification, it was possible to predict the response of the same neurone to step-modulation of intensity at various wavelengths successfully with the first-order Wiener kernel. This demonstrates that the cells behaved linearly under our stimulus conditions, which justifies the use of the first-order kernel as a means to characterize the system we wished to study.

Response latency of brisk-sustained (X) and brisk-transient (Y) cells in the cat retina

1. Several methods for evaluating light-evoked response latency and its variability in brisk-sustained (X) and brisk-transient (Y) retinal ganglion cells were tested. The most accurate procedure proved to be that described by Levick (1973), in which the time of the occurrence of the fourth impulse after stimulus onset is taken as an estimate of the latency.2. The shortest response latencies are obtained when the stimuli are the same size as the receptive field centre. At medium and high response amplitudes (> 150 impulses/sec) the response of brisk-transient (Y) cells to these optimal stimuli is 10-15 msec faster than that of adjacent brisk-sustained (X) cells.3. The response latency of brisk-sustained (X) cells for stimuli larger than the receptive field centre increases, whereas that of brisk-transient (Y) cells remains constant. Brisk-sustained (X) cells respond faster than do brisk-transient (Y) cells to stimuli smaller than the receptive field centre.4. No systematic difference exists between brisk-sustained (X) and brisk-transient (Y) cells in regard to the temporal variability of the response. The standard deviation of the latency for stimuli of optimal size decreases from 2.0-8.0 msec at medium stimulus contrast to 0.6-2.0 msec at high stimulus contrast.5. The response of OFF-centre cells to the disappearance of a light spot is always slower than that of an ON-centre cell of the same class to the onset of this stimulus. However, when OFF-centre cells are stimulated with dark spots, their response latency does not differ from that of ON-centre cells of the same class.6. No simple relationship exists between the response latency and the response amplitude. At medium and high discharge rates, most brisk-transient (Y) cells respond faster than an adjacent brisk-sustained (X) cell with equal response. At the same response amplitude, the latencies become shorter as the background illumination is raised. The same discharge rate can be obtained with stimuli of sub-optimal and supra-optimal size, but the latency for the larger stimulus is shorter than that for the smaller one. Latency, therefore, is an additional parameter characterizing the light-evoked response.

Characterization of spatial and temporal properties of monkey LGN Y-cells

LGN Y-cells in 3 anaesthetized (N2O/O2) and paralyzed rhesus monkeys were investigated with stimuli, intensity modulated by gaussian white noise, and with moving and counterphase modulated spatial sine wave gratings. The results support the model, postulated on the base of electrophysiological recordings in the retina cat and mudpuppy, which consists of a linear centre and surround mechanism whose responses are modified in a frequency-selective multiplicative way by a nonlinear mechanism in the receptive field. This nonlinear mechanism is also held responsible for the second-order harmonic responses, which are the defining characteristic of Y-cells. The temporal and spatial characteristics of these mechanisms were determined. The responses obtained with the GWN stimulation and with modulated spatial sine wave gratings both indicate that the optimal temporal frequency of the linear mechanisms is near 7 Hz and 70 td and near 5 Hz for the nonlinear mechanism. The optimal spatial frequency for the linear mechanism is between 0.5--2 cycles/deg and between 6--12 cycles/deg for the nonlinear mechanism.

A method of nonlinear analysis in the frequency domain

A method is developed for the analysis of nonlinear biological systems based on an input temporal signal that consists of a sum of a large number of sinusoids. Nonlinear properties of the system are manifest by responses at harmonics and intermodulation frequencies of the input frequencies. The frequency kernels derived from these nonlinear responses are similar to the Fourier transforms of the Wiener kernels. Guidelines for the choice of useful input frequency sets, and examples satisfying these guidelines, are given. A practical algorithm for varying the relative phases of the input sinusoids to separate high-order interactions is presented. The utility of this technique is demonstrated with data obtained from a cat retinal ganglion cell of the Y type. For a high spatial frequency grafting, the entire response is contained in the even-order nonlinear components. Even at low contrast, fourth-order components are detectable. This suggests the presence of an essential nonlinearity in the functional pathway of the Y cell, with its singularity at zero contrast.

Protan-like spectral sensitivity of foveal Y ganglion cells of the retina of macaque monkeys

1. The spectral sensitivity of two varieties of macaque Y ganglion cells with a centre-surround organization, type III (non-colour opponent) and type IV (broad-band colour opponent), was examined with test stimuli of different size, shape and wave-length. 2. The spectral sensitivity of type III cells to large stimuli decreased at the long wave-lengths with decreasing retinal eccentricity; this change was due to a lower sensitivity of green-sensitive than of red-sensitive cone input to the surround of foveal cells, which resulted in stronger surround antagonism at the long than at the short wave-lengths leading to a rudimentary form of colour opponency. 3. The spectral properties of foveal type III cells were intermediate between those of perifoveal type III cells, whose surrounds receive a rather similar input from both cone types, and of the predominantly foveal type IV cells, whose surrounds appeared to lack input from green-sensitive cones. 4. The results indicate that both cell types represent varieties within a continuum of a single macaque Y-cell system which has a reduced long-wave-length sensitivity in the foveal region. The fact that a similar reduction of long-wave-length sensitivity can be observed in (foveal) macaque photopic luminosity functions measured with different techniques by different authors suggest that both types of Y cell have an important role in the processing of luminance information.

Enchancement of luminance flicker by color-opponent mechanisms

Color-opponent ganglion cells in the monkey retina respond to luminance flicker at high temporal frequencies. Color opponency, which makes these cells so selective of wavelength at low temporal frequencies, is progressively lost at high frequencies. This loss is due to a frequency-dependent phase shift between the responses of spectrally different center and surround mechanisms in the receptive field of each of these cells. Center and surround responses, which are antagonistic at low temporal frequencies, become synergistic at high ones, making these cells most responsive at high frequencies to those wavelengths to which they are least responsive at low frequencies. This phenomenon can explain the differences between chromatic and luminance flicker in human vision.

Pattern and flicker detection analysed by subthreshold summation

1. We confirm Keesey's (1972) observation that, when a flickering line is viewed, there are distinct thresholds for detecting flicker (or movement) and for detecting a well localized line (pattern detection). Our measurements of the temporal sensitivity of these two mechanisms are similar to Keesey's. 2. The flicker and pattern detection mechanism have been analysed using subthreshold summation, i.e. by observing the effect of subthreshold flickering stimuli (lines and gratings) on the contrast threshold for a flickering test line. 3. The pattern detector shows linear spatial summation of contrast while the flicker detector is non-linear in this respect. 4. The receptive field of the (most sensitive) flicker detector is about two to four times broader than that of the pattern detector. 5. The flicker detector has relatively weak surround inhibition and so, unlike the pattern detector, it is sensitive to a uniform flickering field. 6. The spatial arrangement of the pattern detector is the same at all temporal frequencies (including steady presentation); for flicker detection, the width of the receptive field increases with temporal frequency and the strength of lateral inhibition decreases at high frequencies. 7. Flicker detectors of various widths were demonstrated by using different test stimuli (for 12 Hz modulation); surround ingibition was relatively weak for the broadest detector. 8. There is a delay of surround inhibition of about 3 ms for both flicker and pattern detection. 9. By using a broad test stimulus modulated at a high frequency, a detector can be found with no significant surround inhibition. At threshold, this stimulus produces a sensation of flicker without the appearance of lateral motion observed for finer test lines at lower frequencies. 10. The characteristics of pattern and flicker (movement) detection are compared to electrophysiological studies on X (sustained) and Y (transient) neurones respectively, and correlations are described for studies of temporal frequency response, non-linearity, width of receptive field, strength of the inhibitory surround and motion sensitivity.

The control of retinal ganglion cell discharge by receptive field surrounds

1. This paper describes the behaviour of the receptive field surround, and how surround signals combine with those from the centre to generate the discharge of the retinal ganglion cells of the cat. 2. A small test spot is flashed upon the middle of the receptive field of an on-centre X-cell, alone, or together with a concentric annulus of fixed luminance. The reduction in discharge brought about by the annulus is independent of spot luminance. From this it is inferred that centre and surround signals combine additively. 3. Knowing that the combination of signals is additive, the surround signal can be estimated by comparing the ganglion cell's response to diffuse illumination of its receptive field with that to an equiluminous spot which optimally stimulates the centre while encroaching minimally upon the periphery. 4. Application of this technique to X-cells shows that although the surround seems to have a threshold, it is at its most sensitive in the dark-adapted eye, and typically is only 0.3-0.5 log units less sensitive than the centre. 5. Centre and surround sensitivities are decreased from their dark-adapted levels by increasing background illumination, but the decline of surround sensitivity is initially less rapid than that of the centre. Thus with increasing light-adaptation the surround becomes relatively more sensitive. In the light-adapted eye centre and surround are about equally sensitive to diffuse illumination. 6. Although, in the dark-adapted eye, illumination of the receptive field periphery of an on-centre unit depresses firing, removal of that illumination produces no off-discharge. Off-discharges appear only when background illumination exceeds about 104 quanta (507)/deg 2 sec. This confirms Barlow & Levick (1969b). 7. In the dark-adapted eye surround latency is longer than that of the centre. With increasing background illumination the latency difference is reduced. 8. For X-cells, the rate of the maintained discharge depends to some extent on the balance of centre-surround antagonism. But this antagonism is not the major factor accounting for the relative constancy of mean rate at high background luminances, for the rate then can be almost independent of the size of a steady pot. 9. The mean rate of discharge of Y-cells seems to depend even less upon the balance of centre-surround antagonism. 10. Y-cell surrounds could not properly be isolated with the optimal spot-diffuse illumination technique, so detailed measurements of their behaviour were not made. However, the dark-adapted surround appear to be as sensitive as those of X-cells.

Surround contribution to light adaptation in cat retinal ganglion cells

1. The sensitivity of a cat's retinal ganglion cell to a small, dim, spot flashed upon the middle of the receptive field depends upon the size of a concentric steady background: sensitivity is reduced monotonically with background area. All backgrounds which equal or exceed in size the central summing area of the ganglion cell produce an equivalent reduction of sensitivity, even though only backgrounds which extend outside the central summing area depress the maintained discharge. 2. If a small background lies upon the middle of the receptive field, and the test spot is made intense enough to evoke a strong response, steady illumination of the periphery may make the response larger. 3. This change in response is not due to an enhancement of centre sensitivity by the surround, but is readily understood if steady illumination of the periphery adapts out the surround's antagonism of the centre's response to the test flash. 4. The failure of steady stimulation of the surround to alter centre sensitivity implies that signals from the surround subtract from, or add to, those from the centre.

Dark adaptation and short-wavelength backgrounds decrease perceived size

The effects of background luminance, contrast, and background wavelength on the perceived size of small line figures were studied at mesopic levels of light adaptation. Perceived size diminished at low levels of background luminance. The effect disappeared at high levels of luminance. Perceived size of luminous circles increased as a logarithmic function of background luminance when the background intensity did not exceed 25 td(1). The strength of the size effect decreased as a function of circle diameter from 0-125 to 2 deg of visual angle(2). Perceived size of small luminous circles, subtending less than 0-5 deg, also increased as a function of contrast at low values of contrast but at very high values of contrast there was a decrease in perceived size. Background luminance had the same effect on the perceived size of circles as on the perceived size of spatial cycles in gratings. Control experiments led to the conclusion that dark adaptation is the primary source of the size effects. The main evidence for this conclusion was obtained from a demonstration that the same background luminance produced either an increase or a decrease in perceived size, depending on the adaptational state of the eye. It was also found that a shift from cone vision to rod vision contributes to the effects, for a stimulus looked smaller on a short-wavelength background than on a long-wavelength background. The size effects can be predicted from the changes of receptive-field properties of single neurones under corresponding conditions of stimulation, if it is assumed that the perception of size is mediated by size-specific channels formed of single neurones. Stimulation that leads to an activation of small receptive fields appears to indicate to the brain the presence of small retinal images. If small receptive fields are experimentally made responsive to larger retinal images, an underestimation of size results.

Sensitization and centre-surround antagonism in Necturus retina

1. The impulse discharge of ganglion cells and the proximal negative response were recorded with extracellular micro-electrodes in Necturus retina and spatial influences of background illumination studied.2. When a steady background of constant illuminance was extended in diameter from 0.25 to 1 mm, the response to a small concentrically placed test flash was enhanced (sensitization).3. This sensitization was tonic, showed marked spatial summation, was also found by surrounding a central background with an annulus, and when quantified by threshold measurements, closely resembled the sensitization effect found psychophysically in human vision.4. Sensitization was prominent in intracellular recordings from bipolar cells and usually apparent in horizontal cells. In bipolar cells, a background change which produced sensitization evoked a large but predominantly transient potential whose polarity was opposite to that evoked by a central flash.5. The tonicity and large spatial extent of the sensitization effect strongly suggest that it is mediated by horizontal cells. They might do this by either acting upon receptors, bipolar cells, or both.

Frequency characteristics of retinal neurons in the carp

Frequency characteristics of various retinal neurons in the carp were studied using sinusoidally modulated light as an input. They were affected by both intensity and pattern of illumination. In the horizontal cells, in which the effect of light intensity was studied most extensively, an increase in the light intensity brought about a decrease of the gain, which was more marked at lower frequencies, resulting in a shift of cutoff frequency towards higher frequencies and in a slight low frequency attenuation. A decrease in the area illuminated had an effect similar to a decrease in the light intensity. In the receptor, the low frequency attenuation was not apparent even at high light intensities. The adaptation process in receptors was not sufficient to explain the low frequency attenuation in the horizontal cells, and a possible contribution of negative feedback from horizontal cells to receptors was suggested. In the bipolar cell, the lateral interaction played an important role. An increase in an area resulted in the suppression of the response at low frequencies where the phases of the center and the surround responses were opposed, but in the augmentation near 5 Hz where the two responses were in phase. In amacrine cells, a low frequency attenuation and a phase advance at low frequencies were very prominent, and were considered to be due mainly to a process designated here as the neural adaptation.

Distributed relaxation processes in sensory adaptation

Dynamic description of most receptors, even in their near-linear ranges, has not led to understanding of the underlying physical events-in many instances because their curious transfer functions are not found in the usual repertoire of integral-order control-system analysis. We have described some methods, borrowed from other fields, which allow one to map any linear frequency response onto a putative weighting over an ensemble of simpler relaxation processes. One can then ask whether the resultant weighting of such processes suggests a corresponding plausible distribution of values for an appropriate physical variable within the sensory transducer. To illustrate this approach, we have chosen the fractional-order low-frequency response of Limulus lateral-eye photoreceptors. We show first that the current "adapting-bump" hypothesis for the generator potential can be formulated in terms of local first-order relaxation processes in which local light flux, the cross section of rhodopsin for photon capture, and restoration rate of local conductance-changing capability play specific roles. A representative spatial distribution for one of these parameters, which just accounts for the low-frequency response of the receptor, is then derived and its relation to cellular properties and recent experiments is examined. Finally, we show that for such a system, nonintegral-order dynamics are equivalent to nonhyperbolic statics, and that the efficacy distribution derived to account for the small-signal dynamics in fact predicts several decades of near-logarithmic response in the steady state. Encouraged by the result that one plausible proposal can account approximately for both the low-frequency dynamics (the transfer function s(k)) and the range-compressing statics (the Weber-Fechner relationship) measured in this photoreceptor, we have described some formally similar applications of these distributed effects to the vertebrate retina and to analogous properties of mechanoreceptors and chemoreceptors.

Adaptation and dynamics of cat retinal ganglion cells

1. The impulse/quantum (I/Q) ratio was measured as a function of background illumination for rod-dominated, pure central, linear square-wave responses of retinal ganglion cells in the cat.2. The I/Q ratio was constant at low backgrounds (dark adapted state) and inversely proportional to the 0.9 power of the background at high backgrounds (the light adapted state). There was an abrupt transition from the dark-adapted state to the light-adapted state.3. It was possible to define the adaptation level at a particular background as the ratio (I/Q ratio at that background)/(dark adapted I/Q ratio).4. The time course of the square-wave response was correlated with the adaptation level. The response was sustained in the dark-adapted state, partially transient at the transition level, and progressively more transient the lower the impulse/quantum ratio of the ganglion cell became. This was true both for on-centre and off-centre cells.5. The frequency response of the central response mechanism at different adaptation levels was measured. It was a low-pass characteristic in the dark-adapted state and became progressively more of a bandpass characteristic as the cell became more light-adapted.6. The rapidity of onset of adaptation was measured with a time-varying adapting light. The impulse/quantum ratio is reset within 100 msec of the onset of the conditioning light, and is kept at the new value throughout the time the conditioning light is on.7. These results can be explained by a nonlinear feedback model. In the model, it is postulated that the exponential function of the horizontal cell potential controls transmission from rods to bipolars. This model has an abrupt transition from dark- to light-adapted states, and its response dynamics are correlated with adaptation level.

Luminance-dependent changes in mesopic visual contrast sensitivity

Spatial and temporal modulation transfer functions have been measured as a function of luminance at scotopic and mesopic levels. It is found that throughout the scotopic range the data can be accounted for by a two process visual system, with the processes having the following properties. Simple excitatory processThis mechanism summates over somewhat less than 1 square degree of visual angle and over about 200 msec of time. These properties do not change with luminance, but the over-all sensitivity of the mechanism follows the De Vries-Rose law. The temporal properties of this mechanism follows those of the scotopic b-wave. At high luminances, the temporal, but not spatial, properties of this mechanism break down in a manner which had not been studied.Low-frequency inhibitory processThis process is manifest as a decrease in sensitivity from that of the simple excitatory process. Its effect is observed only when both spatial and temporal frequency are low, suggesting that it has a larger (perhaps 3 x) summation area in both space and time. The inhibitory process is not observed below a certain threshold luminance, which is highly dependent upon the configuration of the stimulus. For a suitable stimulus, this threshold will be well into the scotopic luminance range.It is suggested that these two processes represent psychophysical correlates of the centre and surround of retinal receptive fields.

Properties of the surround response mechanism of cat retinal ganglion cells and centre-surround interaction

1. The properties of the surround response mechanism of on-centre cells and its interaction with the centre mechanism were studied by recording from single optic tract fibres. In many of the experiments the spatial distribution of the light within the retinal image of the stimuli was measured.2. Pure surround responses of on-centre cells were isolated using a centrally located steady light which selectively desensitized (adapted) the centre mechanism. This permitted a peripheral flashing stimulus whose luminance varied over a range as great as 1.38 log units to elicit surround responses which, for any given cell, were of invariant shape. The rate of decay of the firing frequency of the spike burst at ;off' varied from cell to cell. The general characteristics of such pure surround responses to squarewave stimuli were described. The plot of the magnitude of pure responses against stimulus luminance, at constant background conditions, was curvilinear.3. The pure surround response of two off-centre cells was isolated; it was similar in shape to the pure centre response of on-centre cells.4. Interaction of centre and surround mechanisms of on-centre cells was studied by eliciting a pure central and a pure surround response from the same cell. The electronically obtained algebraic sum of these two pure responses equalled the mixed response of the ganglion cell to simultaneous presentation of the stimuli which evoked the pure responses when presented singly. This is probably best explained by algebraic summation of centre and surround inputs.5. The pure surround response from two cells to a fixed flashing stimulus was attenuated by a steady field adapting light, both when this was superimposed upon the stimulus and when not superimposed. In the latter case, (i) when the spatial separation between the flashing stimulus and the adapting light was at a minimum, less than 10% of the adapting flux fell inside the boundaries of the stimulating flux, and (ii) the response was attenuated also if the adapting light was in the geometric centre of the receptive field. These results indicate that the adaptation pool of the surround mechanism extends to the central portions of the receptive field.6. Nearly half the cells tested did not yield pure surround responses. This was probably due to differences, within the ganglion cell population, (i) of the spatial distribution of the ratio of centre to surround signal sensitivity and (ii) of differences in the ratio of centre to surround adaptivity in the receptive field middle. It was not due to excess adaptive flux falling outside the region of maximal centre mechanism adaptivity, nor due to excess stimulus flux falling inside the region of maximal signal sensitivity.

Dynamic characteristics of retinal ganglion cell responses in goldfish

A cross-correlation technique has been applied to quantify the dependence of the dynamic characteristics of retinal ganglion cell responses in goldfish on intensity, wavelength, spatial configuration, and spot size. Both theoretical and experimental evidence justify the use of the cross-correlation procedure which allows the completion of rather extensive measurements in a relatively short time. The findings indicate the following. (a) The shape of the amplitude characteristics depends on the energy per unit of time (power) falling within the center of a receptive field rather than on the intensity of the stimulus spot. For spot diameters of up to 1 mm, identical amplitude characteristics can be obtained by interchanging area and intensity. Therefore the receptor processes do not contribute to the change in the amplitude characteristics as a function of the power of the stimulus light. (b) For high frequencies the amplitude characteristics obtained as a function of power join together in a common envelope if plotted on an absolute sensitivity scale. For spontaneous ganglion cells this envelope holds over a range of three log units and the shape is identical for central and peripheral processes. (c) The amplitude characteristics of the central and peripheral processes converging to a ganglion cell are identical, irrespective of the sign (on or off) and the spectral coding of the response. Therefore we have no evidence for interneurons in the goldfish retina unique to the periphery of the receptive field.