The contribution of central and peripheral vision in scene categorization: a study on people with central vision loss

The importance of peripheral vision when searching 3D real-world scenes: A gaze-contingent study in virtual reality

Visual search in natural scenes is a complex task relying on peripheral vision to detect potential targets and central vision to verify them. The segregation of the visual fields has been particularly established by on-screen experiments. We conducted a gaze-contingent experiment in virtual reality in order to test how the perceived roles of central and peripheral visions translated to more natural settings. The use of everyday scenes in virtual reality allowed us to study visual attention by implementing a fairly ecological protocol that cannot be implemented in the real world. Central or peripheral vision was masked during visual search, with target objects selected according to scene semantic rules. Analyzing the resulting search behavior, we found that target objects that were not spatially constrained to a probable location within the scene impacted search measures negatively. Our results diverge from on-screen studies in that search performances were only slightly affected by central vision loss. In particular, a central mask did not impact verification times when the target was grammatically constrained to an anchor object. Our findings demonstrates that the role of central vision (up to 6 degrees of eccentricities) in identifying objects in natural scenes seems to be minor, while the role of peripheral preprocessing of targets in immersive real-world searches may have been underestimated by on-screen experiments.

Effects of Transient Loss of Vision on Head and Eye Movements during Visual Search in a Virtual Environment

Central and peripheral fields of view extract information of different quality and serve different roles during visual tasks. Past research has studied this dichotomy on-screen in conditions remote from natural situations where the scene would be omnidirectional and the entire field of view could be of use. In this study, we had participants looking for objects in simulated everyday rooms in virtual reality. By implementing a gaze-contingent protocol we masked central or peripheral vision (masks of 6 deg. of radius) during trials. We analyzed the impact of vision loss on visuo-motor variables related to fixation (duration) and saccades (amplitude and relative directions). An important novelty is that we segregated eye, head and the general gaze movements in our analyses. Additionally, we studied these measures after separating trials into two search phases (scanning and verification). Our results generally replicate past on-screen literature and teach about the role of eye and head movements. We showed that the scanning phase is dominated by short fixations and long saccades to explore, and the verification phase by long fixations and short saccades to analyze. One finding indicates that eye movements are strongly driven by visual stimulation, while head movements serve a higher behavioral goal of exploring omnidirectional scenes. Moreover, losing central vision has a smaller impact than reported on-screen, hinting at the importance of peripheral scene processing for visual search with an extended field of view. Our findings provide more information concerning how knowledge gathered on-screen may transfer to more natural conditions, and attest to the experimental usefulness of eye tracking in virtual reality.

Different hemispheric specialization for face/word recognition: A high-density ERP study with hemifield visual stimulation

INTRODUCTION

The right fusiform face area (FFA) is important for face recognition, whereas the left visual word fusiform area (VWFA) is critical for word processing. Nevertheless, the early stages of unconscious and conscious face and word processing have not been studied systematically.

MATERIALS AND METHODS

To explore hemispheric differences for face and word recognition, we manipulated the visual field (left vs. right) and stimulus duration (subliminal [17 ms] versus supraliminal [300 ms]). We recorded P100 and N170 peaks with high-density ERPs in response to faces/objects or Japanese words/scrambled words in 18 healthy young subjects.

RESULTS

Contralateral P100 was larger than ipsilateral P100 for all stimulus types in the supraliminal, but not subliminal condition. The face- and word-N170s were not evoked in the subliminal condition. The N170 amplitude for the supraliminal face stimuli was significantly larger than that for the objects, and right hemispheric specialization was found for face recognition, irrespective of stimulus visual hemifield. Conversely, the supraliminal word-N170 amplitude was not significantly modulated by stimulus type, visual field, or hemisphere.

CONCLUSIONS

These results suggest that visual awareness is crucial for face and word recognition. Our study using hemifield stimulus presentation further demonstrates the robust right FFA for face recognition but not the left VWFA for word recognition in the Japanese brain.

Emergence of Visual Center-Periphery Spatial Organization in Deep Convolutional Neural Networks

Research at the intersection of computer vision and neuroscience has revealed hierarchical correspondence between layers of deep convolutional neural networks (DCNNs) and cascade of regions along human ventral visual cortex. Recently, studies have uncovered emergence of human interpretable concepts within DCNNs layers trained to identify visual objects and scenes. Here, we asked whether an artificial neural network (with convolutional structure) trained for visual categorization would demonstrate spatial correspondences with human brain regions showing central/peripheral biases. Using representational similarity analysis, we compared activations of convolutional layers of a DCNN trained for object and scene categorization with neural representations in human brain visual regions. Results reveal a brain-like topographical organization in the layers of the DCNN, such that activations of layer-units with central-bias were associated with brain regions with foveal tendencies (e.g. fusiform gyrus), and activations of layer-units with selectivity for image backgrounds were associated with cortical regions showing peripheral preference (e.g. parahippocampal cortex). The emergence of a categorical topographical correspondence between DCNNs and brain regions suggests these models are a good approximation of the perceptual representation generated by biological neural networks.

Object search in neovascular age-related macular degeneration: the crowding effect

BACKGROUND

Visual search, an activity that relies on central vision, is frequent in daily life. This study investigates the effect of spacing between items in an object search task in participants with central vision loss.

METHODS

Patients with neovascular age-related macular degeneration (AMD), age-matched controls, and young controls were included. The stimuli were displays of four, six and nine objects randomly presented in a 'crowded' (spacing 1.5°) or 'uncrowded' (spacing 6°) condition. For each of 96 trials, participants were asked to search for a predefined target that remained on the screen until the response was recorded. Accuracy, search time, and eye movements (number of fixations and scan path ratio) were recorded.

RESULTS

Compared to older controls, accuracy decreased by 31 per cent and search time increased by 61 per cent in AMD participants. Ageing also affected performance with a lower accuracy by 13.5 per cent and longer search times by 46 per cent in older compared to younger controls. Increasing the spacing between elements increased accuracy by 21 per cent in AMD participants but it had no effect in older and younger controls. Performance was not related to visual acuity or to duration of neovascular AMD, but search time was correlated to the lesion size in the 'crowded' condition.

CONCLUSIONS

Object search is ubiquitous in daily life activities. When visual acuity is irrevocably reduced, increasing the spacing between elements can reliably improve object search performance in patients.

Age-related macular degeneration changes the processing of visual scenes in the brain

In age-related macular degeneration (AMD), the processing of fine details in a visual scene, based on a high spatial frequency processing, is impaired, while the processing of global shapes, based on a low spatial frequency processing, is relatively well preserved. The present fMRI study aimed to investigate the residual abilities and functional brain changes of spatial frequency processing in visual scenes in AMD patients. AMD patients and normally sighted elderly participants performed a categorization task using large black and white photographs of scenes (indoors vs. outdoors) filtered in low and high spatial frequencies, and nonfiltered. The study also explored the effect of luminance contrast on the processing of high spatial frequencies. The contrast across scenes was either unmodified or equalized using a root-mean-square contrast normalization in order to increase contrast in high-pass filtered scenes. Performance was lower for high-pass filtered scenes than for low-pass and nonfiltered scenes, for both AMD patients and controls. The deficit for processing high spatial frequencies was more pronounced in AMD patients than in controls and was associated with lower activity for patients than controls not only in the occipital areas dedicated to central and peripheral visual fields but also in a distant cerebral region specialized for scene perception, the parahippocampal place area. Increasing the contrast improved the processing of high spatial frequency content and spurred activation of the occipital cortex for AMD patients. These findings may lead to new perspectives for rehabilitation procedures for AMD patients.

Primary visual cortical remapping in patients with inherited peripheral retinal degeneration

Human studies addressing the long-term effects of peripheral retinal degeneration on visual cortical function and structure are scarce. Here we investigated this question in patients with Retinitis Pigmentosa (RP), a genetic condition leading to peripheral visual degeneration. We acquired functional and anatomical magnetic resonance data from thirteen patients with different levels of visual loss and twenty-two healthy participants to study primary (V1) visual cortical retinotopic remapping and cortical thickness. We identified systematic visual field remapping in the absence of structural changes in the primary visual cortex of RP patients. Remapping consisted in a retinotopic eccentricity shift of central retinal inputs to more peripheral locations in V1. Importantly, this was associated with changes in visual experience, as assessed by the extent of the visual loss, with more constricted visual fields resulting in larger remapping. This pattern of remapping is consistent with expansion or shifting of neuronal receptive fields into the cortical regions with reduced retinal input. These data provide evidence for functional changes in V1 that are dependent on the magnitude of peripheral visual loss in RP, which may be explained by rapid cortical adaptation mechanisms or long-term cortical reorganization. This study highlights the importance of analyzing the retinal determinants of brain functional and structural alterations for future visual restoration approaches.

How does age-related macular degeneration affect real-world visual ability and quality of life? A systematic review

OBJECTIVES

To review systematically the evidence of age-related macular degeneration (AMD) affecting real-world visual ability and quality of life (QoL). To explore trends in specific topics within this body of the literature.

DESIGN

Systematic review.

METHODS

A systematic literature search was carried out using MEDLINE, EMBASE, CINAHL, PsycINFO, PsychARTICLES and Health and Psychosocial Instruments for articles published up to January 2015 for studies including people diagnosed with AMD, assessing real-world visual ability or QoL as an outcome. Two researchers screened studies for eligibility. Details of eligible studies including study design, characteristics of study population and outcomes measured were recorded in a data extraction table. All included studies underwent quality appraisal using the Mixed Methods Appraisal Tool 2011 Version (MMAT).

RESULTS

From 5284 studies, 123 were eligible for inclusion. A range of approaches were identified, including performance-based methods, quantitative and qualitative patient-reported outcome measures (PROMs). AMD negatively affects tasks including mobility, face recognition, perception of scenes, computer use, meal preparation, shopping, cleaning, watching TV, reading, driving and, in some cases, self-care. There is evidence for higher rates of depression among people with AMD than among community dwelling elderly. A number of adaptation strategies have been associated with AMD of varying duration. Much of the research fails to report the type of AMD studied (59% of included studies) or the duration of disease in participants (74%). Of those that do report type studied, the breakdown is as follows: wet AMD 20%, dry AMD 4% and both types 17%.

CONCLUSIONS

There are many publications highlighting the negative effects of AMD in various domains of life. Future research should focus on delivering some of this research knowledge into patient management and clinical trials and differentiating between the types of AMD.

Scene perception in age-related macular degeneration: Effect of spatial frequencies and contrast in residual vision

Age-related macular degeneration (AMD) is characterized by a central vision loss. Here, we investigated the ability of AMD patients to process the spatial frequency content of scenes in their residual vision, depending of the luminance contrast level. AMD patients and normally-sighted elderly participants (controls) performed a categorization task involving large scenes (outdoors vs. indoors) filtered in low spatial frequencies (LSF), high spatial frequencies (HSF), and non-filtered scenes (NF). Luminance contrast of scenes was equalized between stimuli using a root-mean square (RMS) contrast normalization. In Experiment 1, we applied an RMS contrast of 0.1 (for luminance values between 0 and 1), a value situated between the mean contrast of LSF and HSF scenes in natural conditions. In Experiment 2, we applied an RMS contrast of 0.3, corresponding to the mean contrast of HSF scenes in natural conditions. In Experiment 3, we manipulated four levels of linearly-increasing RMS contrasts (0.05, 0.10, 0.15, and 0.20) for HSF scenes only. Compared to controls, AMD patients gave more non-responses in the categorization of HSF than NF or LSF scenes, irrespective of the contrast level of scenes. Performances improved as contrast increased in HSF scenes. Controls were not differentially affected by the spatial frequency content of scenes. Overall, results suggest that LSF processing is well preserved in AMD patients and allows efficient scene categorization in their parafoveal residual vision. The HSF processing deficit could be partially restored by enhancing luminance contrast.

Categorization Task over a Touch Screen in Age-Related Macular Degeneration

PURPOSE

In our modern society, many touch screen applications require hand-eye coordination to associate an icon with its specific contextual unit on phones, on computers, or in public transport. We assessed the ability of patients with age-related macular degeneration (AMD) to explore scenes and to associate a target (animal or object) with a unique congruent scene (e.g., to match a fish with the sea) presented between three other distractors on a touch screen computer.

METHODS

Twenty-four patients with AMD (64 to 90 years) with best-corrected visual acuity between 20/40 and 20/400 as well as 17 age-matched (60 to 94 years) and 15 young (22 to 34 years) participants with normal visual acuity had to match a target with a congruent scene by moving their index finger on a 22-in touch screen.

RESULTS

Patients were as accurate (98.7% correct responses) as the age-matched control (98.9% correct responses) and young participants (99.3% correct responses) at performing the task. The duration of exploration was significantly longer for the AMD patients (mean, 4.13 seconds) compared with the age-matched group (mean, 2.96 seconds). The young participants were also significantly faster than the old group (mean, 0.93 seconds). The movement parameters of the older participants (patients and old control subjects) were affected compared with the young; the peak speed decreased (-8 cm/s) and the movement duration increased (+0.9 seconds) with age compared with the young group.

CONCLUSIONS

People with AMD are able to perform a contextual association task on a touch screen with high accuracy. The AMD patients were specifically affected in the "exploration" phase; their accuracy and movement parameters did not differ from the old control group. Our study suggests that the decline associated with AMD is more focused on the duration of exploration than on movement parameters in touch screen use.