Choroidalyzer: An Open-Source, End-to-End Pipeline for Choroidal Analysis in Optical Coherence Tomography.

Purpose: To develop Choroidalyzer, an open-source, end-to-end pipeline for segmenting the choroid region, vessels, and fovea, and deriving choroidal thickness, area, and vascular index. Methods: We used 5600 OCT B-scans (233 subjects, six systemic disease cohorts, three device types, two manufacturers). To generate region and vessel ground-truths, we used state-of-the-art automatic methods following manual correction of inaccurate segmentations, with foveal positions manually annotated. We trained a U-Net deep learning model to detect the region, vessels, and fovea to calculate choroid thickness, area, and vascular index in a fovea-centered region of interest. We analyzed segmentation agreement (AUC, Dice) and choroid metrics agreement (Pearson, Spearman, mean absolute error [MAE]) in internal and external test sets. We compared Choroidalyzer to two manual graders on a small subset of external test images and examined cases of high error. Results: Choroidalyzer took 0.299 seconds per image on a standard laptop and achieved excellent region (Dice: internal 0.9789, external 0.9749), very good vessel segmentation performance (Dice: internal 0.8817, external 0.8703), and excellent fovea location prediction (MAE: internal 3.9 pixels, external 3.4 pixels). For thickness, area, and vascular index, Pearson correlations were 0.9754, 0.9815, and 0.8285 (internal)/0.9831, 0.9779, 0.7948 (external), respectively (all P < 0.0001). Choroidalyzer's agreement with graders was comparable to the intergrader agreement across all metrics. Conclusions: Choroidalyzer is an open-source, end-to-end pipeline that accurately segments the choroid and reliably extracts thickness, area, and vascular index. Especially choroidal vessel segmentation is a difficult and subjective task, and fully automatic methods like Choroidalyzer could provide objectivity and standardization.

Collective plasticity of binocular interactions in the adult visual system.

Binocular visual plasticity can be initiated via either bottom-up or top-down mechanisms, but it is unknown if these two forms of adult plasticity can be independently combined. In seven participants with normal binocular vision, sensory eye dominance was assessed using a binocular rivalry task, before and after a period of monocular deprivation and with and without selective attention directed towards one eye. On each trial, participants reported the dominant monocular target and the inter-ocular contrast difference between the stimuli was systematically altered to obtain estimates of ocular dominance. We found that both monocular light- and pattern-deprivation shifted dominance in favour of the deprived eye. However, this shift was completely counteracted if the non-deprived eye's stimulus was selectively attended. These results reveal that shifts in ocular dominance, driven by bottom-up and top-down selection, appear to act independently to regulate the relative contrast gain between the two eyes.

Assessing the reliability of web-based measurements of visual function.

Many behavioural phenomena have been replicated using web-based experiments, but evaluation of the agreement between objective measures of web- and lab-based performance is required if scientists and clinicians are to reap the benefits of web-based testing. In this study, we investigated the reliability of a task which assesses early visual cortical function by evaluating the well-known 'oblique effect' (we are better at seeing horizontal and vertical edges than tilted ones) and the levels of agreement between remote, web-based measures and lab-based measures. Sixty-nine young participants (mean age, 21.8 years) performed temporal and spatial versions of a web-based, two-alternative forced choice (2AFC) orientation-identification task. In each case, orientation-identification thresholds (the minimum orientation difference at which a standard orientation could be reliably distinguished from a rotated comparison) were measured for cardinal (horizontal and vertical) and oblique orientations. Reliability was assessed in a subsample of 18 participants who performed the same tasks under laboratory conditions. Robust oblique effects were found, such that thresholds were substantially lower for cardinal orientations compared to obliques, for both web- and lab-based measures of the temporal and spatial 2AFC tasks. Crucially, web- and lab-based orientation-identification thresholds showed high levels of agreement, demonstrating the suitability of web-based testing for assessments of early visual cortical function. Future studies should assess the reliability of similar web-based tasks in clinical populations to evaluate their adoption into clinical settings, either to screen for visual anomalies or to assess changes in performance associated with progression of disease severity.

Eye movements in visual impairment.

This Special Issue describes the impact of visual impairment on visuomotor function. It includes contributions that examine gaze control in conditions associated with abnormal visual development such as amblyopia, dyslexia and neurofibromatosis as well as disorders associated with field loss later in life, such as macular degeneration and stroke. Specifically, the papers address both gaze holding (fixation), and gaze-following behavior (single saccades, sequences of saccades and smooth-pursuit) that characterize active vision in daily life and evaluate the influence of both pathological and simulated field loss. Several papers address the challenges to reading and visual search; describing how the patterns of eye movements in these real-world tasks adapt to visual impairment and highlighting how they could serve as diagnostic markers of visuomotor function.

The effects of simulated hemianopia on eye movements during text reading.

Vision loss is a common, devastating complication of cerebral strokes. In some cases the complete contra-lesional visual field is affected, leading to problems with routine tasks and, notably, the ability to read. Although visual information crucial for reading is imaged on the foveal region, readers often extract useful parafoveal information from the next word or two in the text. In hemianopic field loss, parafoveal processing is compromised, shrinking the visual span and resulting in slower reading speeds. Recent approaches to rehabilitation using perceptual training have been able to demonstrate some recovery of useful visual capacity. As gains in visual sensitivity were most pronounced at the border of the scotoma, it may be possible to use training to restore some of the lost visual span for reading. As restitutive approaches often involve prolonged training sessions, it would be beneficial to know how much recovery is required to restore reading ability. To address this issue, we employed a gaze-contingent paradigm using a low-pass filter to blur one side of the text, functionally simulating a visual field defect. The degree of blurring acts as a proxy for visual function recovery that could arise from restitutive strategies, and allows us to evaluate and quantify the degree of visual recovery required to support normal reading fluency in patients. Because reading ability changes with age, we recruited a group of younger participants, and another with older participants who are closer in age to risk groups for ischaemic strokes. Our results show that changes in patterns of eye movement observed in hemianopic loss can be captured using this simulated reading environment. This opens up the possibility of using participants with normal visual function to help identify the most promising strategies for ameliorating hemianopic loss, before translation to patient groups.

The spatial properties of adaptation-induced distance compression.

Exposure to a dynamic texture reduces the perceived separation between objects, altering the mapping between physical relations in the environment and their neural representations. Here we investigated the spatial tuning and spatial frame of reference of this aftereffect to understand the stage(s) of processing where adaptation-induced changes occur. In Experiment 1, we measured apparent separation at different positions relative to the adapted area, revealing a strong but tightly tuned compression effect. We next tested the spatial frame of reference of the effect, either by introducing a gaze shift between adaptation and test phase (Experiment 2) or by decoupling the spatial selectivity of adaptation in retinotopic and world-centered coordinates (Experiment 3). Results across the two experiments indicated that both retinotopic and world-centered adaptation effects can occur independently. Spatial attention to the location of the adaptor alone could not account for the world-centered transfer we observed, and retinotopic adaptation did not transfer to world-centered coordinates after a saccade (Experiment 4). Finally, we found that aftereffects in different reference frames have a similar, narrow spatial tuning profile (Experiment 5). Together, our results suggest that the neural representation of local separation resides early in the visual cortex, but it can also be modulated by activity in higher visual areas.

Disruption of Positional Encoding at Small Separations in the Amblyopic Periphery.

Purpose: Positional judgments in amblyopia are impaired more at the center of the visual field than in the periphery. However, the effects of visual field position frequently are confounded with stimulus separation. The purpose of this experiment was to parse the effects of stimulus separation and eccentricity on the positional deficit in amblyopia. Methods: Subjects adjusted the positions of stimuli of varying separations on isoeccentric arcs. The task was simultaneous bisection and alignment of broadband, high-contrast, uncrowded targets with reference to central fixation. Ten strabismic amblyopes and five normally sighted controls performed the task dichoptically; a subset of amblyopes performed the task monocularly with the amblyopic eye. Spread (inverse of precision) and bias were measured at multiple visual field locations comprising two to three separation \(\times\) four eccentricity conditions in each visual field quadrant. Results: In normal controls, both spread and bias increased with eccentricity, and spread (but not bias) increased linearly with separation until 7° eccentricity. Strabismic amblyopes showed a different profile: spread and bias were higher at small separations at all eccentricities, such that performance showed a quadratic trend against separation. Thus, at each eccentricity, the difference in performance between groups was largest at the smallest separation. Conclusions: These results are consistent with disruptions in Weber mechanisms of positional encoding in strabismic amblyopia, and indicate that binocular stimulation by proximal targets produces a loss of spatial precision well beyond the fovea.

Attentional eye selection modulates sensory eye dominance.

Brief periods of monocular deprivation significantly modify binocular visual processing. For example, patching one eye for a few hours alters the inter-ocular balance, with the previously patched eye becoming dominant once the patch is removed. However, the contribution of higher-level visual processing to this phenomenon is still unclear. Here, we compared changes in sensory eye dominance produced by three types of monocular manipulations in adult participants with normal binocular vision. One eye was covered for 150 min using either an opaque patch, a diffusing lens, or a prism that inverted the image. All three manipulations altered dominance duration and predominance during binocular rivalry (BR) in favour of the treated eye and the time courses of the changes were similar. These results indicate that modifications of luminance or contrast are not strictly necessary to drive shifts in eye dominance, as both were unaltered in the prism condition. Next, we found that shifts in eye dominance were dependent on attentional demands during the monocular treatment period, providing support for the role of attentional eye selection in modulating eye dominance. Finally, we found relatively rapid build-up of the ocular dominance shift after the onset of monocular treatment. Taken together, our results suggest that modifications to monocular input alter inter-ocular balance via selective attentional mechanisms that bias output towards the deprived eye. Eye-based attention may play an important role in conditions where normal input to one eye is disrupted, such as childhood amblyopia.

Learning to silence saccadic suppression.

Perceptual stability is facilitated by a decrease in visual sensitivity during rapid eye movements, called saccadic suppression. While a large body of evidence demonstrates that saccadic programming is plastic, little is known about whether the perceptual consequences of saccades can be modified. Here, we demonstrate that saccadic suppression is attenuated during learning on a standard visual detection-in-noise task, to the point that it is effectively silenced. Across a period of 7 days, 44 participants were trained to detect brief, low-contrast stimuli embedded within dynamic noise, while eye position was tracked. Although instructed to fixate, participants regularly made small fixational saccades. Data were accumulated over a large number of trials, allowing us to assess changes in performance as a function of the temporal proximity of stimuli and saccades. This analysis revealed that improvements in sensitivity over the training period were accompanied by a systematic change in the impact of saccades on performance-robust saccadic suppression on day 1 declined gradually over subsequent days until its magnitude became indistinguishable from zero. This silencing of suppression was not explained by learning-related changes in saccade characteristics and generalized to an untrained retinal location and stimulus orientation. Suppression was restored when learned stimulus timing was perturbed, consistent with the operation of a mechanism that temporarily reduces or eliminates saccadic suppression, but only when it is behaviorally advantageous to do so. Our results indicate that learning can circumvent saccadic suppression to improve performance, without compromising its functional benefits in other viewing contexts.

Impact of microsaccades on visual shape processing.

Sensitivity to subtle changes in the shape of visual objects has been attributed to the existence of global pooling mechanisms that integrate local form information across space. Although global pooling is typically demonstrated under steady fixation, other work suggests prolonged fixation can lead to a collapse of global structure. Here, we ask whether small ballistic eye movements that naturally occur during periods of fixation affect the global processing of radial frequency (RF) patterns-closed contours created by sinusoidally modulating the radius of a circle. Observers were asked to discriminate the shapes of circular patterns and RF-modulated patterns while fixational eye movements were recorded binocularly at 500 Hz. Microsaccades were detected using a velocity-based algorithm, allowing trials to be sorted according to the relative timing of stimulus and microsaccade onset. Results revealed clear perisaccadic changes in shape discrimination thresholds. Performance was impaired when microsaccades occurred close to stimulus onset, but facilitated when they occurred shortly afterward. In contrast, global integration of shape was unaffected by the timing of microsaccades. These findings suggest that microsaccades alter the discrimination sensitivity to briefly presented shapes but do not disrupt the spatial pooling of local form signals.NEW & NOTEWORTHY Microsaccades cause rapid displacement of visual images during fixation and dramatically alter the perception of basic image features. However, their effect on more complex aspects of visual processing is not well understood. Here, we demonstrate a dissociation in the impact of microsaccades on shape perception. Although overall shape discrimination performance is modulated around the time of microsaccades, the pooling efficiency of global mechanisms that combine local form information across space remains unaffected.

Linking Multi-Modal MRI to Clinical Measures of Visual Field Loss After Stroke.

Loss of vision across large parts of the visual field is a common and devastating complication of cerebral strokes. In the clinic, this loss is quantified by measuring the sensitivity threshold across the field of vision using static perimetry. These methods rely on the ability of the patient to report the presence of lights in particular locations. While perimetry provides important information about the intactness of the visual field, the approach has some shortcomings. For example, it cannot distinguish where in the visual pathway the key processing deficit is located. In contrast, brain imaging can provide important information about anatomy, connectivity, and function of the visual pathway following stroke. In particular, functional magnetic resonance imaging (fMRI) and analysis of population receptive fields (pRF) can reveal mismatches between clinical perimetry and maps of cortical areas that still respond to visual stimuli after stroke. Here, we demonstrate how information from different brain imaging modalities-visual field maps derived from fMRI, lesion definitions from anatomical scans, and white matter tracts from diffusion weighted MRI data-provides a more complete picture of vision loss. For any given location in the visual field, the combination of anatomical and functional information can help identify whether vision loss is due to absence of gray matter tissue or likely due to white matter disconnection from other cortical areas. We present a combined imaging acquisition and visual stimulus protocol, together with a description of the analysis methodology, and apply it to datasets from four stroke survivors with homonymous field loss (two with hemianopia, two with quadrantanopia). For researchers trying to understand recovery of vision after stroke and clinicians seeking to stratify patients into different treatment pathways, this approach combines multiple, convergent sources of data to characterize the extent of the stroke damage. We show that such an approach gives a more comprehensive measure of residual visual capacity-in two particular respects: which locations in the visual field should be targeted and what kind of visual attributes are most suited for rehabilitation.

Does physical exercise and congruent visual stimulation enhance perceptual learning?

PURPOSE: There is currently great interest in methods that can modulate brain plasticity, both in terms of understanding the basic mechanisms, and in the remedial application to situations of sensory loss. Recent work has focussed on how different manipulations might be combined to produce new settings that reveal synergistic actions. Here we ask whether a prominent example of adult visual plasticity, called perceptual learning, is modified by other environmental factors, such as visual stimulation and physical exercise. METHODS: We quantified the magnitude, rate and transfer of perceptual learning using a peripheral Vernier alignment task, in two groups of subjects matched for a range of baseline factors (e.g. age, starting Vernier threshold, baseline fitness). We trained subjects for 5 days on a Vernier alignment task. In one group, we introduced an exercise protocol with congruent visual stimulation. The control group received the same visual stimulation, but did not exercise prior to measurement of Vernier thresholds. RESULTS: Although the task generated large amounts of learning (~40%) and some transfer to untrained conditions in both groups, there were no specific benefits associated with either the addition of an exercise schedule or congruent visual stimulation. CONCLUSION: In adults, short periods of physical exercise and visual stimulation do not enhance perceptual learning.

Publisher Correction: Temporal rate is not a distinct perceptual metric.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

The Effect of Perceptual Learning on Face Recognition in Individuals with Central Vision Loss.

Purpose: To examine whether perceptual learning can improve face discrimination and recognition in older adults with central vision loss. Methods: Ten participants with age-related macular degeneration (ARMD) received 5 days of training on a face discrimination task (mean age, 78 ± 10 years). We measured the magnitude of improvements (i.e., a reduction in threshold size at which faces were able to be discriminated) and whether they generalized to an untrained face recognition task. Measurements of visual acuity, fixation stability, and preferred retinal locus were taken before and after training to contextualize learning-related effects. The performance of the ARMD training group was compared to nine untrained age-matched controls (8 = ARMD, 1 = juvenile macular degeneration; mean age, 77 ± 10 years). Results: Perceptual learning on the face discrimination task reduced the threshold size for face discrimination performance in the trained group, with a mean change (SD) of -32.7% (+15.9%). The threshold for performance on the face recognition task was also reduced, with a mean change (SD) of -22.4% (+2.31%). These changes were independent of changes in visual acuity, fixation stability, or preferred retinal locus. Untrained participants showed no statistically significant reduction in threshold size for face discrimination, with a mean change (SD) of -8.3% (+10.1%), or face recognition, with a mean change (SD) of +2.36% (-5.12%). Conclusions: This study shows that face discrimination and recognition can be reliably improved in ARMD using perceptual learning. The benefits point to considerable perceptual plasticity in higher-level cortical areas involved in face-processing. This novel finding highlights that a key visual difficulty in those suffering from ARMD is readily amenable to rehabilitation.

Short-term monocular deprivation reduces inter-ocular suppression of the deprived eye.

The adult visual system was traditionally thought to be relatively hard-wired, but recent studies have challenged this view by demonstrating plasticity following short-term monocular deprivation. Depriving one eye of spatial information for 2-3 h increased subsequent sensory dominance of that eye. However, the mechanism underlying this phenomenon is unclear. The present study sought to address this issue and determine the consequences of short-term monocular deprivation on inter-ocular suppression of each eye. Sensory eye dominance was examined before and after depriving an eye of all input using an opaque patch for 2.5 h, in six adult participants with normal binocular vision. We used a percept tracking task during binocular rivalry (BR) to assess the relative eye dominance, and an objective probe detection task under continuous flash suppression (CFS) to quantify each eye's susceptibility to inter-ocular suppression. The monocular contrast increment threshold of each eye was also measured using the probe task to ascertain if the altered eye dominance is accompanied by changes in monocular perception. Our BR results replicated previous findings of a shift of relative dominance towards the eye that has been deprived of form information. More crucially, using CFS we demonstrated reduced inter-ocular suppression of the deprived eye with no complementary changes in the other eye, and no monocular changes in increment threshold. These findings imply that short-term monocular deprivation alters binocular interactions. The differential effect on inter-ocular suppression between eyes may have important implications for the use of patching as a therapy to recover visual function in amblyopia.

Temporal rate is not a distinct perceptual metric.

Sensory adaptation experiments have revealed the existence of 'rate after-effects' - adapting to a relatively fast rate makes an intermediate test rate feel slow, and adapting to a slow rate makes the same moderate test rate feel fast. The present work aims to deconstruct the concept of rate and clarify how exactly the brain processes a regular sequence of sensory signals. We ask whether rate forms a distinct perceptual metric, or whether it is simply the perceptual aggregate of the intervals between its component signals. Subjects were exposed to auditory or visual temporal rates (a 'slow' rate of 1.5 Hz and a 'fast' rate of 6 Hz), before being tested with single unfilled intervals of varying durations. Results show adapting to a given rate strongly influences the perceived duration of a single empty interval. This effect is robust across both interval reproduction and duration discrimination judgments. These findings challenge our understanding of rate perception. Specifically, they suggest that contrary to some previous assertions, the perception of sequence rate is strongly influenced by the perception of the sequence's component duration intervals.

Temporal modulation improves dynamic peripheral acuity.

Macular degeneration and related visual disorders greatly limit foveal function, resulting in reliance on the peripheral retina for tasks requiring fine spatial vision. Here we investigate stimulus manipulations intended to maximize peripheral acuity for dynamic targets. Acuity was measured using a single interval orientation discrimination task at 10° eccentricity. Two types of image motion were investigated along with two different forms of temporal manipulation. Smooth object motion was generated by translating targets along an isoeccentric path at a constant speed (0-20°/s). Ocular motion was simulated by jittering target location using previously recorded fixational eye movement data, amplified by a variable gain factor (0-8). In one stimulus manipulation, the sequence was temporally subsampled by displaying the target on an evenly spaced subset of video frames. In the other, the contrast polarity of the stimulus was reversed at a variable rate. We found that threshold under object motion was improved at all speeds by reversing contrast polarity, while temporal subsampling improved resolution at high speeds but impaired performance at low speeds. With simulated ocular motion, thresholds were consistently improved by contrast polarity reversal, but impaired by temporal subsampling. We find that contrast polarity reversal and temporal subsampling produce differential effects on peripheral acuity. Applying contrast polarity reversal may offer a relatively simple image manipulation that could enhance visual performance in individuals with central vision loss.

Individual variation in inter-ocular suppression and sensory eye dominance.

The competitive and inhibitory interactions between the two eyes' images are a pervasive aspect of binocular vision. Over the last decade, our understanding of the neural processes underpinning binocular rivalry (BR) and continuous flash suppression (CFS) has increased substantially, but we still have little understanding of the relationship between these two effects and their variation in the general population. Studies that pool data across individuals and eyes risk masking substantial variations in binocular vision that exist in the general population. To investigate this issue we compared the depth of inter-ocular suppression evoked by BR with that elicited by CFS, in a group (N = 25) of visually normal individuals. A noise pattern (either static for BR or dynamic for CFS) was presented to one eye and its suppressive influence on a probe grating presented simultaneously to the other eye was measured. We found substantial individual differences in the magnitude of suppression (a 10-fold variation in probe detection threshold) evoked by each task, but performance on BR was a significant predictor of performance on the CFS task. However many individuals showed marked asymmetries between the two eyes' ability to detect a suppressed target, that were not necessarily the same for the two tasks. There was a tendency for the magnitude of the asymmetry to increase as the refresh rate of the dynamic noise increased. The results suggest a common underlying mechanism is likely to be responsible, at least in part, for driving inter-ocular suppression under BR and CFS. The marked asymmetries in inter-ocular suppression at higher noise refresh rates, may be indicative of a difference in temporal processing between the eyes.

Clinical investigation plan for the use of interactive binocular treatment (I-BiT) for the management of anisometropic, strabismic and mixed amblyopia in children aged 3.5-12 years: a randomised controlled trial.

BACKGROUND: Amblyopia (lazy eye) affects the vision of approximately 2% of all children. Traditional treatment consists of wearing a patch over their 'good' eye for a number of hours daily, over several months. This treatment is unpopular and compliance is often low. Therefore, results can be poor. I-BiT is a system, based on stereo technology using shutter glasses, designed to treat amblyopia using dichoptic stimulation. This trial uses a redesigned system for home use and includes eye-tracking capability. METHODS/DESIGN: This is a randomised controlled trial involving three groups of 40 patients each, aged between 3.5 and 12 years, with a diagnosis of (1) anisometropic amblyopia, (2) mixed or strabismic amblyopia prior to strabismic surgery and (3) mixed or strabismic amblyopia who have just undergone strabismus surgery. They will be randomised in a 1:1 ratio between I-BiT and control and will receive treatment, at home over a 6-week period. Their visual acuity will be assessed independently at baseline, mid-treatment (week 3), at the end of treatment (week 6) and, for those receiving the active I-BiT treatment, 4 weeks after completing treatment (week 10). The primary endpoint will be the change in visual acuity from baseline to the end of treatment. Secondary endpoints will be additional visual acuity measures, patient acceptability, compliance and the incidence of adverse events. DISCUSSION: This is a randomised controlled trial using the redesigned I-BiT™ system to determine if this is a feasible treatment strategy for the management of anisometropic, strabismic and mixed amblyopia. TRIAL REGISTRATION: ISRCTN Number/Clinical trials.gov, ID: NCT02810847 . Registered on 23 June 2016.