John G. Robson

Spatial-Frequency Channels in Human Vision*

Psychometric functions were determined concurrently for detection of simple gratings (luminance sinusoidally modulated with spatial frequency f) and complex gratings (luminance modulated by the sum of two sinusoids, with frequencies f and f′). Results were used to test the hypothesis that the two components of a complex grating may be detected independently. In an extensive experiment with f = 14 cycles/deg, the independence hypothesis was consistently rejected only when f/f′=54 or 45, but rarely rejected when the value of f/f′ lay outside this range. In other experiments, f was between 1.9 and 22.4 cycles/deg. All results are compatible with the assumption that the human visual system contains sensory channels, each selectively sensitive to different narrow ranges of spatial frequencies, whose outputs are detected independently.

Discrimination at threshold: labelled detectors in human vision.

Abstract We examined discriminations between small patches of grating that differed in either spatial or temporal frequency. The patches were presented at contrasts near to detection threshold. For certain pairs of stimuli, each was correctly identified as often as it was detected. To explain this result, we hypothesize that the detectors of these stimuli arelabelled, in the sense that the observer can distinguish the response of each detector from that of any other. Under this assumption, we find that the detectors form two non-overlapping sets in their selectivity for temporal frequency. In their selectivity for spatial frequency, the detectors of slowly varying stimuli can be partitioned into 7 distinct sets, but only 3 sets are evident among the detectors of rapidly modulated patterns.

Fluctuations of accommodation under steady viewing conditions

It is well known that under steady environmental conditions motor systems exhibit residual fluctuations or unrest. The tremor in skeletal muscles has been extensively studied and shows a dominant frequency component of 10 c/s (Schaefer, 1886). The pupil of the eye shows a physiological unrest with a high frequency component of 1F2 c/s (Stark, Campbell & Atwood, 1958). The tremor of the eyeball during steady fixation has been described as having dominant 30-80 c/s components (Ditchburn & Ginsborg, 1953). While investigating the characteristics of the accommodation response of the eye we noticed that the refractive power undergoes small fluctuations and it is the purpose of this paper to describe these. Attempts to account for motor tremor have often led to feedback theory being invoked (Hammond, Merton & Sutton, 1956; Lippold, Redfearn & Vuco, 1957). The implication of the feedback concept in this connexion is that sensory information from, for example, muscle spindles is fed back to the motor control centre in order that a desired tension or length response of the muscle shall be achieved or maintained (Granit, 1955). In the accommodation system the sensory information comes from the retina and here we have the advantage of easy optical access by natural means, enabling us to manipulate independently the sensory information sent back to the controlling nerve centre, a facility not readily available in other motor systems. Quantitative as well as qualitative data about its components and their interaction are necessary before a feedback system can be fully described and quantitative predictions made. In this paper we are reporting measurements that must be integrated into a quantitative description of the accommodation system if it is given in terms of feedback theory.

Grating summation in fovea and periphery

Abstract Results from previous studies measuring the detectability of sinusoidal gratings have been interpreted by models postulating several sizes of receptive fields. It has not been clear, however, whether or not these several sizes coexist at a single position in the visual field. Perhaps there is only one size centered at each position, but the size varies as a function of eccentricity. In this study, the detectability of compound gratings containing two sinusoidal components was compared to that of each component alone. Measurements were made in the fovea and 7.5° into the periphery. Stimuli were localized in a small region of the visual field and sharp spatial and temporal transients eliminated by weighting grating contrast with Gaussian functions of space and time. To reduce possible effects of expectation, bias and frequency uncertainty, a temporal, forced-choice, interlaced staircase procedure was used. The results are consistent with models postulating several sizes of receptive fields at each position in the visual field but not with models postulating only one size at each position, even when the size varies as a function of eccentricity to account for the differences in spatial interaction characteristic of different parts of the visual field.

Spatio-temporal interactions in cat retinal ganglion cells showing linear spatial summation.

The spatio‐temporal characteristics of cat retinal ganglion cells showing linear summation have been studied by measuring both magnitude and phase of the responses of these cells to drifting or sinusoidally contrast‐modulated sinusoidal grating patterns. It has been demonstrated not only that X cells behave approximately linearly when responding with amplitudes of less than about 10 impulses/sec to stimuli of low contrast but also that cells of another type with larger receptive field centres (Q cells) behave approximately linearly under the same conditions. These Q cells appear to form a homogeneous group which is probably a subset of the tonic W cells (Stone & Fukuda, 1974) or sluggish centre‐surround cells (Cleland & Levick, 1974). The over‐all spatio‐temporal frequency characteristics of cells showing linear spatial summation are not separable in space and time. The form of the spatial frequency responsivity function of these cells depends upon the temporal frequency at which it is measured while the temporal phase of their resonse measured at any constant temporal frequency depends upon the spatial frequency of the stimulus. The behaviour of X and Q cells is quite well explained by an extension of the model in which signals from centre and surround mechanisms with radially Gaussian weighting functions are summed to provide the drive to the retinal ganglion cell. While the general form of the temporal frequency response characteristics of these ganglion cells are probably provided by the characteristics of elements common to the centre and surround pathways, the spatio‐temporal interactions can be explained by assuming that the surround signal is delayed relative to the centre signal by a few milliseconds.

The scotopic threshold response of the dark-adapted electroretinogram of the mouse

The most sensitive response in the dark‐adapted electroretinogram (ERG), the scotopic threshold response (STR) which originates from the proximal retina, has been identified in several mammals including humans, but previously not in the mouse. The current study established the presence and assessed the nature of the mouse STR. ERGs were recorded from adult wild‐type C57/BL6 mice anaesthetized with ketamine (70 mg kg−1) and xylazine (7 mg kg−1). Recordings were between DTL fibres placed under contact lenses on the two eyes. Monocular test stimuli were brief flashes (λmax 462 nm; ‐6.1 to +1.8 log scotopic Troland seconds(sc td s)) under fully dark‐adapted conditions and in the presence of steady adapting backgrounds (‐3.2 to ‐1.7 log sc td). For the weakest test stimuli, ERGs consisted of a slow negative potential maximal ≈200 ms after the flash, with a small positive potential preceding it. The negative wave resembled the STR of other species. As intensity was increased, the negative potential saturated but the positive potential (maximal ≈110 ms) continued to grow as the b‐wave. For stimuli that saturated the b‐wave, the a‐wave emerged. For stimulus strengths up to those at which the a‐wave emerged, ERG amplitudes measured at fixed times after the flash (110 and 200 ms) were fitted with a model assuming an initially linear rise of response amplitude with intensity, followed by saturation of five components of declining sensitivity: a negative STR (nSTR), a positive STR (pSTR), a positive scotopic response (pSR), PII (the bipolar cell component) and PIII (the photoreceptor component). The nSTR and pSTR were approximately 3 times more sensitive than the pSR, which was approximately 7 times more sensitive than PII. The sensitive positive components dominated the b‐wave up to > 5 % of its saturated amplitude. Pharmacological agents that suppress proximal retinal activity (e.g. GABA) minimized the pSTR, nSTR and pSR, essentially isolating PII which rose linearly with intensity before showing hyperbolic saturation. The nSTR, pSTR and pSR were desensitized by weaker backgrounds than those desensitizing PII. In conclusion, ERG components of proximal retinal origin that are more sensitive to test flashes and adapting backgrounds than PII provide the ‘threshold’ negative and positive (b‐wave) responses of the mouse dark‐adapted ERG. These results support the use of the mouse ERG in studies of proximal retinal function.

Dissecting the dark-adapted electroretinogram

Although gross recordings of the ganzfeld flash-evoked electroretinogram (ERG) can potentially provide information about the activity of many, if not all, retinal cell types, it is necessary to dissect the ERG into its components to realize this potential fully. Here we describe various procedures that have been used in intact mammalian eyes to identify and characterize the contributions to the dark-adapted ERG of different cells in the retinal rod pathway. These include (1) examination of the very early part of the response to a flash (believed to reflect directly the photocurrent of rods), (2) application of high-energy probe flashes to provide information about the underlying rod photoreceptor response even when this component is obscured by the responses of other cells, (3) pharmacological suppression of responses of amacrine and ganglion cells to identify the contribution of these cells and to reveal the weaker responses of bipolar cells, (4) use of pharmacological agents that block transmission of signals from rods to more proximal neurons to separate responses of rods from those of later neurons, (5) examination of the ERG changes produced by ganglion-cell degeneration or pharmacological block of nerve-spike generation to identify the contribution of spiking neurons, (6) modeling measured amplitude-energy functions and timecourse of flash responses and (7) using steady backgrounds to obtain differential reductions in sensitivity of different cell types. While some of these procedures can be applied to humans, the results described here have all been obtained in studies of the ERG of anaesthetized cats, or macaque monkeys whose retinas are very similar to those of humans.

Rod and cone contributions to the a‐wave of the electroretinogram of the macaque

The electroretinogram (ERG) of anaesthetised dark‐adapted macaque monkeys was recorded in response to ganzfeld stimulation and rod‐ and cone‐driven receptoral and postreceptoral components were separated and modelled. The test stimuli were brief (< 4.1 ms) flashes. The cone‐driven component was isolated by delivering the stimulus shortly after a rod‐saturating background had been extinguished. The rod‐driven component was derived by subtracting the cone‐driven component from the mixed rod–cone ERG. The initial part of the leading edge of the rod‐driven a‐wave scaled linearly with stimulus energy when energy was sufficiently low and, for times less than about 12 ms after the stimulus, it was well described by a linear model incorporating a distributed delay and three cascaded low‐pass filter elements. Addition of a simple static saturating non‐linearity with a characteristic intermediate between a hyperbolic and an exponential function was sufficient to extend application of the model to most of the leading edge of the saturated responses to high energy stimuli. It was not necessary to assume involvement of any other non‐linearity or that any significant low‐pass filter followed the non‐linear stage of the model. A negative inner‐retinal component contributed to the later part of the rod‐driven a‐wave. After suppressing this component by blocking ionotropic glutamate receptors, the entire a‐wave up to the time of the first zero‐crossing scaled with stimulus energy and was well described by summing the response of the rod model with that of a model describing the leading edge of the rod‐bipolar cell response. The negative inner‐retinal component essentially cancelled the early part of the rod‐bipolar cell component and, for stimuli of moderate energy, made it appear that the photoreceptor current was the only significant component of the leading edge of the a‐wave. The leading edge of the cone‐driven a‐wave included a slow phase that continued up to the peak, and was reduced in amplitude either by a rod‐suppressing background or by the glutamate analogue, cis‐piperidine‐2,3‐dicarboxylic acid (PDA). Thus the slow phase represents a postreceptoral component present in addition to a fast component of the a‐wave generated by the cones themselves. At high stimulus energies, it appeared less than 5 ms after the stimulus. The leading edge of the cone‐driven a‐wave was adequately modelled as the sum of the output of a cone photoreceptor model similar to that for rods and a postreceptoral signal obtained by a single integration of the cone output. In addition, the output of the static non‐linear stage in the cone model was subject to a low‐pass filter with a time constant of no more than 1 ms. In conclusion, postreceptoral components must be taken into account when interpreting the leading edge of the rod‐ and cone‐driven a‐waves of the dark‐adapted ERG.

Identifying Inner Retinal Contributions to the Human Multifocal ERG

Contributions to the multifocal electroretinogram (ERG) from the inner retina (i.e. ganglion and amacrine cells) were identified by recording from monkeys before and after intravitreal injections of n-methyl DL aspartate (NMDLA) and/or tetrodotoxin (TTX). Components similar in waveform to those removed by the drugs were identified in the human multifocal ERG if the stimulus contrast was set at 50% rather than the typically employed 100% contrast. These components were found to be missing or diminished in the records from some patients with glaucoma and diabetes, diseases which affect the inner retina.

Evidence for a ganglion cell contribution to the primate electroretinogram (ERG): effects of TTX on the multifocal ERG in macaque.

To assess the contribution of spiking inner retinal neurons to the multifocal electroretinogram (ERG), recordings were made from four monkeys (Macaca mulatta) before and after intravitreal injections of tetrodotoxin (TTX). TTX blocks all sodium-based action potentials and thus terminates spiking activity of amacrine and ganglion cells. TTX eliminated a large component from the control responses, and this TTX-sensitive component was present as early as 10 ms after the stimulus. Before injection with TTX, the 103 focal ERG responses varied in waveform across the retina. After TTX, the response waveforms were largely independent of retinal position, indicating that it was primarily the TTX-sensitive component of the control response that was dependent upon retinal location. Given that retinal ganglion cells compose a sizable proportion of the retinal elements that produce action potentials, it is likely that part of the TTX-sensitive component is due to the spiking activity of these cells. Further, the systematic change in waveform of the TTX-sensitive component with distance from the optic nerve head suggests that part of the TTX-sensitive component may originate from the activity of the ganglion cell axons. Based on these findings, there is reason to be optimistic that the multifocal technique can be employed to study the effects of glaucoma and other diseases that affect the inner retina.