Stereoblind Subjects Research Articles

Corrigendum: The Influence of the Stimulus Design on the Harmonic Components of the Steady-State Visual Evoked Potential

[This corrects the article DOI: 10.3389/fnhum.2020.00343.].

Assessing stereo blindness and stereo acuity on digital displays

Previous studies on stereoscopic acuity have shown that the percentage of stereo blind subjects is relevant. Moreover, stereoscopic visualization is becoming widely diffused in different fields, like, e.g., entertainment, surgery or VR training, where it is necessary an accurate assessment of stereoscopic abilities of the involved subjects. Therefore, there might be the need of performing a stereo blindness and stereo acuity test before each visualization session involving stereoscopic images. In this paper, we propose a method to assess stereo acuity and stereo blindness directly on the chosen device, under the same visualization condition and setup adopted for the tasks to perform, in order to have the same perceptual response. We present software-based tests suitable for a generic stereoscopic displays, and we compare their effectiveness performing a comparison with a standard physical, card-based, test commonly used in assessment of stereo acuity and stereo blindness. We provide to the reader all the details to perform autonomously the tests, of which images will be downloadable from web.

Depth Perception, Binocular Integration and Hand-Eye Coordination in Intact and Stereo Impaired Human Subjects

A series of tests was devised to assess stereoscopic depth processing, motion parallax depth processing, binocular integration and hand-eye coordination in normal, stereoblind, and stereo deficient subjects. Using a random- dot stereoscopic display viewed through a stereoscope we established that of the 262 subjects tested, 177 were categorized as having normal stereoscopic depth perception, 28 as being stereo deficient and 57 as being stereoblind. These three groups of subjects processed motion parallax information for depth equally well, but the stereoblind and stereo deficient subjects had significantly longer reaction times. On the hand-eye coordination tests the stereoblind subjects performed significantly less well than did normal and stereo deficient subjects, whose performance on these tests was similar. Our binocular integration tests revealed significantly less integration in stereoblind subjects than in normal and stereo deficient subjects. The tests we have devised will be useful for the accurate assessment of various forms of treatment for amblyopia and strabismus for the reinstatement of depth perception, hand-eye coordination and binocular integration.

Better performance with two eyes than with one in stereo-blind subjects’ judgments of motion in depth

Visual tasks that involve judging distance or depth obviously benefit from considering retinal disparities and ocular convergence, but various simple visual thresholds are also lower when looking with two eyes. This is also true for stereo-blind subjects. One benefit of using two eyes is that looking with two eyes provides two chances of making the critical distinction. From the literature it would appear that using two eyes might only be an advantage for low contrast stimuli and simple tasks. We here demonstrate that stereo-blind subjects can benefit from using two eyes when making judgments about clearly visible complex stimuli. The task was to judge the direction of rotation of a simulated transparent cylinder. Stereo-blind subjects performed better when looking with two eyes than when looking with their preferred eye. It did not matter for their performance whether the images in their two eyes were correlated or not. Various control experiments ascertained that they judged the direction of rotation from the images in each eye separately and then combined these judgments, rather than relying on differences between the images in the two eyes. These findings raise doubts about the validity of using monocular vision as a control for quantitative studies of the use of binocular disparity.

Texture discrimination and visual search: development, learning and plasticity

The abilities to segment coherent surfaces out of a discrepant background (texture segmentation) and to localize single objects out of a complex scene (visual search) are believed to occur at a peripheral level in the cortical visual pathway. Are texture segmentation and visual search innate or acquired? Can these abilities be improved in adulthood by learning? Does a disturbed early visual experience (strabismus, amblyopia) lead to a reduction of these abilities? Texture segmentation and visual search undergo a protracted development which continues well into adulthood. Even in adult observers, these abilities can be improved with practice. Learning is enduring, but not specific: It can be transferred from one task to another and between the two eyes of a given subject (including the two eyes of stereoblind subjects). We conclude that learning of texture segmentation and visual search does not involve the specific features of a visual scene, but rather reflects an improved strategy of localizing visual objects in a cluttered environment. This type of learning probably occurs at a very high level of cortical processing.

Dichoptic luminance beat visual evoked potentials in the assessment of binocularity in children.

Direct evidence of a distinct cortical binocular pathway has been provided by the production of nonlinear (difference) beats from dichoptic luminance stimulation in stereonormal adults and the absence or diminution of these beats in stereoblind subjects. We have investigated a clinically useful application of this technique in a pediatric population with potentially abnormal binocular vision. We recorded dichoptic luminance beat visual evoked potentials (VEPs) from 20 children (ages 7 months to 8 years) with abnormal binocular ability secondary to strabismus and/or amblyopia and compared this to a control group of 20 children with normal binocularity. Stereoblind children generated significantly lower dichoptic signal-to-noise ratios than stereonormal children (P < .001). Responses to monoptic multifrequency flicker were not significantly different between the two groups (P = .936). This dichoptic VEP can be performed quickly and easily on young children and gives a quantitative assessment of cortical binocularity that may not be determinable by standard clinical methods. This technique may also prove useful for the preoperative gradation of binocular potential and prediction of postoperative binocular fusion.

The Pulfrich pendulum phenomenon in stereoblind subjects.

The Pulfrich pendulum phenomenon, in which a pendulum swinging in the frontoparallel plane appears to swing in an ellipse when a neutral density filter is placed over one eye of the observer, was investigated in stereoblind subjects. It was found that such subjects can report the presence of the Pulfrich effect although they fail to fuse random-dot stereograms and fail to exhibit interocular transfer of the movement aftereffect. These findings suggest that 'stereoblind' subjects must retain some residual binocular mechanism for depth perception. Three possibilities are considered: (i) the stereoblind may be able to utilise contiguous temporal disparities as a cue for depth in the Pulfrich effect; (ii) they may retain some residual binocularity, sufficient to reveal the Pulfrich effect but not for other more demanding tasks for the binocular mechanisms; and (iii) they may retain some coarse magnocellular pathway disparity mechanism having lost their high-acuity parvocellular disparity system. There is little evidence to support any of these hypotheses, but the third shows promise.

Binocular beats: Psychophysical studies of binocular interaction in normal and stereoblind humans

We describe a psychophysical method for assessing binocular integration using dichoptically-presented uniform fields. By temporally modulating uniform field luminances at different frequencies between the eyes, a rhythmic beat is produced—a visual percept characterized by undulations in luminance at a frequency equal to the arithmetic difference between the two monocular stimulus frequencies. Using a signal detection paradigm, we studied the beat as a function of modulation depth in normal and binocularly deficient subjects. Normal subjects easily detected the beat, even at low modulation depths, while stereoblind subjects (with stereoacuity worse than 2000 sec arc) failed to detect beats at any modulation depth or with any combination of stimulus frequencies tested. The beat provides evidence for the confluence of monocular signals into binocular integrating mechanisms. Our results therefore suggest that the status of functioning binocular mechanisms is related to the detectability of the beat. This uniform-field stimulus does not require accurate accommodation, fixation, vergence or high spatial resolution, thus making this technique particularly attractive for the study of binocular interaction in developing infants and in binocularly deficient adults.

Dynamic stereoacuity: a comparison of electrophysiological and psychophysical responses in normal and stereoblind observers.

We compared stereoscopic thresholds determined psychophysically and by using visually evoked potentials in normal adult and stereoblind subjects. For normal observers there was a strong linear relationship between visual evoked potential amplitude and disparity (r = 0.86 on average). However, stereoblind observers' responses showed no relationship to disparity (r = 0.18 on average). For normal observers, log visual evoked potential stereo thresholds determined by an extrapolation procedure were on average within 0.17 octaves of the psychophysically determined log thresholds. They did not differ significantly from psychophysical thresholds, and they correlated significantly with behavioral thresholds (r = 0.86).

Stereoscopic contours and optokinetic nystagmus in normal and stereoblind subjects

Moving stereoscopic contours in a dynamic random-dot stereogram have been previously shown to induce optokinetic nystagmus in subjects with normal stereopsis. For this to be validated as an objective test of stereopsis, stereoblind subjects must also be shown not to develop OKN, especially since it has been shown that the optomotor system of stereoblind individuals retains sensitivity to some cyclopean stimuli. In this report we verify that stereoblind subjects do not have an optomotor response to stereoscopic contours—regardless of the alignment angle at which the stereo image pair is presented.

Topographical study of stereo-related potentials.

In order to estimate objectively binocular vision and especially stereopsis, random dot stereograms generated by a personal computer were used. Brain activity during stereopsis was topographically studied by visually evoked potentials (VEPs). The potentials evoked by binocular viewing of patterns without disparity, e.g. correlogram, were very similar to the potentials evoked from patterns with disparity, i.e. stereogram, as many authors have already indicated. To derive the stereo-related potentials from the VEP elicited by stereograms, the potentials evoked by correlograms were subtracted from the potentials evoked by stereograms, and the differences of topographical distribution between normal and stereoblind subjects were investigated.

Simultaneous binocular integration of the visual tilt effect in normal and stereoblind observers

Apparent displacement of an edge from the subjective visual vertical is typically induced by the simultaneous inspection of a large rotating textured background. In the present series of experiments we have tested the binocular integration of such an effect in normal and stereoblind observers in whom inspection of the rotating background and vision of the vertical edge are presented to different eyes. In normal subjects there was no difference between the binocular and the dichoptic viewing condition while in stereoblind subjects there was a clearcut decrease in the strength of the tilt effect under dichoptic conditions. A residual tilt effect, however, was still present in the latter group under dichoptic conditions and this shows that binocular interactions related to visual tilt can be dissociated from those required for stereopsis.

The binocular contribution to monocular optokinetic nystagmus and after nystagmus asymmetries in humans

Stereoblind subjects show marked monocular asymmetries in both OKN and OKAN, while normal subjects produce more balanced nystagmus. There is a negative relationship between the magnitude of the asymmetries and the magnitude of a binocular contribution to OKN production as measured with dichoptic, strobe illumination. A second experiment with normal subjects corroborated this relationship, revealing a small, but significant asymmetry in these subjects' OKAN. Further, there was no correlation between the monocular nystagmus asymmetries and stereoacuity. The results are explained in terms of subcortical binocular mechanisms.

On the relation of stereoacuity to interocular transfer of the motion and the tilt aftereffects

Interocular transfer of both the motion and the tilt aftereffect were tested in 43 subjects with varying degrees of stereopsis, ranging from normal to stereoblind. Although there was an overall tendency for transfer to diminish with decreasing Stereovision, stereoblind subjects always showed some transfer of the tilt aftereffect, and sometimes also of the motion aftereffect, while some subjects with normal stereothresholds had greatly reduced or no interocular transfer. No quantitative correlation between stereothresholds and amount of transfer could be found. The results indicate that there is no simple relationship between interocular transfer, stereopsis and cortical binocularity, as had been suggested previously.

Binocular vision in normal and stereoblind subjects.

This article describes several different psychophysical strategies that have been applied to the study of binocular vision in normal persons and people with deficient stereopsis. Results from these three techniques--binocular summation, interocular transfer, and utrocular discrimination--are considered in terms of the properties of neural mechanisms that integrate inputs from the two eyes.

Visual evoked potentials to binocular stimulation with disparate patterns.

Experiments with VEP to checkerboard stimulation in 4 subjects with normal stereoscopic vision showed that binocular interaction was the same for fused images and for linear pattern displacement induced by horizontal and vertical prisms, as long as the patterns were projected on the macular areas of both eyes. Also in a stereo-blind subject the impaired binocular VEP interaction remained unchanged during linear displacement of the patterns. Thus the previously reported differences in binocular VEP interaction between stereo-normal stereoblind subjects cannot be explained by binocular stimulation with unfused patterns in the latter group. With patterns displaced to different hemifields in the 2 eyes, no binocular interaction was recorded. Experiments with black-and-white gratings of different orientations to the 2 eyes showed that binocular VEP reactions remained the same even with patterns perpendicular to each other. However, the binocular interaction was disrupted when the subject viewed the perpendicular patterns through coloured filters.

A binocular contribution to the production of optokinetic nystagmus in normal and stereoblind subjects

Viewing a large patterned field moving in one direction produces a regular pattern of eye movements known as Optokinetic Nystagmus (OKN). Fox et al. (1978) showed that the mechanism producing OKN can utilize purely binocular or “cyclopean” input. Experiments reported here confirm that finding and extend it to demonstrate that the binocularity of this mechanism is not disrupted in subjects who lack stereopsis. This result indicates that there exists more than one binocular process in the visual system.

Cyclopean stimulation can influence sensations of self-motion in normal and stereoblind subjects.

Subjects experienced an illusion of self-motion when viewing the randomly patterned inner surface of a cylinder rotating about their main body axis. This sensation of rotation in a direction opposite to the direction of cylinder rotation is known as circular vection. An experiment was conducted to ascertain if the production of circular vection involved a binocular process in the visual system. Using dichoptic strobe illumination, stimuli were created that were identical monocularly but different binocularly. Groups of normal and stereoblind subjects were tested. The presence of purely binocular (cyclopean) stimulation increased the reported magnitude of vection for both groups. We conclude that a binocular process is involved in the production of circular vection and that this process retains its binocularity in stereoblind subjects.

Residual binocular interaction in stereoblind humans

Measurements were made of the extent of interocular transfer of the temporary elevation of grating contrast thresholds that follows inspection of high contrast gratings in 4 normal and 7 stereoblind or stereoanomalous individuals. In common with other aftereffects, the normal observers showed partial interocular transfer of this phenomenon. This was also true of all the stereoblind subjects, although the extent of interocular transfer was reduced. This finding stands in striking contrast to the failure of the same subjects to show interocular transfer of the tilt and motion aftereffects. Evidently stereoblind subjects retain a population of binocular neurons that are unable to mediate either stereopsis or interocular transfer of many other aftereffects.

Evidence for separable binocular processes differentially affected by artificially induced anisometropia.

Two measurable effects are produced by viewing a large disk rotating about the line of sight: a perceived displacement of the apparent horizon (tilt) and cyclotorsion of the eye (torsion). Previous results in our laboratory (see Refs.) using dichoptic presentation techniques show that both effects demonstrate binocular enhancement in normal subjects. Further, stereoblind subjects maintain enhancement for the oculomotor effect, torsion, but not for the perceptual effect, tilt. In this study we have found that this stereoblind pattern of results can be produced optically in normals by introducing a three diopter myopia in one eye. This renders the subject stereoblind and causes suppression of the artificially myopic eye. In this condition torsion still shows enhancement; tilt does not. This procedure may be useful in isolating causes of stereoblindness and in locating different sites of binocular interaction.