A common framework for ambylopic, developmental and adult forms of visual crowding

In our visual field, objects are typically easy to see when we look directly at them and difficult to see in the periphery (the 'edges' of our vision). This difference is not simply to do with resolution - even when a target object in the periphery is large enough to be seen in isolation, the placement of other objects in its immediate vicinity can render the target 'jumbled' and impossible to recognise. This is known as 'visual crowding'. It is crowding and not simple resolution that limits our object recognition over more than 95% of the visual field.

Although crowding does not affect the centre of gaze in 'normal' vision, it becomes elevated the case of strabismic amblyopia, often known as 'lazy eye'. Amblyopia is the most common cause of visual impairment in children and affects ~3% of the population. It is defined by impaired resolution in one eye that occurs despite optical correction. In addition, the amblyopic eye is also strongly affected by crowding - even when objects are large enough to be seen in isolation, vision at the centre of gaze is highly impaired in cluttered environments. Recent research suggests that the central vision of children younger than 12 is similarly affected by crowding. That is, where adults can recognise closely spaced objects when gazing directly at them, the same is not true for children. This places a significant restriction on the vision of children and processes such as their ability to learn to read. However, we know very little about it.

We have here three instances of visual crowding: in normal peripheral vision, in amblyopic eyes and in the eyes of young children. What we do not know is whether these three forms of crowding have a common underlying cause. Can we consider each in the same way, using the same theory, and thus apply the same treatments and coping strategies to overcome crowding? Our hypothesis is that crowding in all its forms represents a form of 'compulsory integration' that occurs when the brain is faced with limited resources: when the cells in our brain that process vision are insufficient in number, they pool across extended regions to compensate. The aim of this project is to compare these three forms of crowding and determine whether they fit this common framework.

The research will be conducted by psychologists, neuroscientists, orthoptists and ophthalmologists, with four specific goals. First, we will determine how crowding affects the appearance of objects. That is, do cluttered scenes look the same to children in their central vision as they do to adults in the periphery? This is central to testing the predictions of the compulsory integration theory and will lead to the development of computer models that simulate vision in these circumstances. Second, we will examine whether the peripheral field in normal vision is like the central vision of those with amblyopia - are there differences between the eyes and can we reduce them in the same way that we treat amblyopia? This creates the possibility for the 'normal' periphery to become a testing ground for new treatments of amblyopia, without the need to risk the central vision of children. Third, we will examine whether crowding is a single process in vision, or a general form of integration that affects different objects separately. For instance, does crowding affect moving objects in the same way that it affects letters? This will not only test the compulsory integration hypothesis in each instance of crowding, but also has the potential to develop new, more sensitive tests for crowding. Finally, we will scan the brains of 'normal' and amblyopic adults to measure whether the variation in crowding that occurs across the visual field follows the variation in neural resources to these same parts of the visual field. This will allow a direct test of the compulsory integration hypothesis in normal and amblyopic vision.

Although it is clear that crowding occurs in three instances of vision - the normal periphery, amblyopic fovea and the developing fovea - it is not known whether these three instances arise from the same mechanism. Our aim is to test whether all three can be commonly understood as a 'compulsory integration' of the visual field, using psychophysics, computational modelling and neuroimaging.

There are four research objectives. First, one of the hallmarks of peripheral crowding is that it systematically changes the appearance of objects. We will determine whether this is also true in amblyopic and developmental crowding. By examining what children see under crowded conditions we can develop computational models of the process and compare their operation to models of adult crowding.

Second, the key feature of amblyopic crowding is that it is strong in the amblyopic eye and minimal in the fellow eye. Our pilot testing suggests that there are similar interocular differences in regions of the normal periphery. We will measure these 'amblyopic zones' and examine whether 'treatment programs', designed for amblyopia, can reduce these interocular differences in the periphery.

Third, we will examine whether crowding in each of the three instances occurs in the same way for the different modalities of vision (e.g. motion vs. colour). Crowding varies in strength across both the visual field and individuals, with amblyopia an extreme case, but it is not known whether these variations are common to all modalities. In measuring this, we will assess whether crowding is a singular process or a more distributed mechanism.

Finally, we will directly measure the neural correlates of peripheral and amblyopic crowding using functional neuroimaging. A combination of structural and functional MRI scans will measure the degree of cortical resources devoted to regions of the visual field and relate this to the strength of crowding in both the normal periphery and amblyopia.

The proposed research will deliver impact in a number of key ways. Outside of the academic impacts outlined above, our efforts in understanding the causes of crowding in amblyopia will have a strong influence on clinical vision and the treatment of visual disorders. Amblyopia is one of the most common childhood disorders of vision, affecting ~3% of the population. In addition to its definitive acuity deficit, strabismic amblyopia in particular is also characterised by elevated crowding in the amblyopic eye. However, crowding is rarely highlighted as a specific therapeutic target. We suspect that significant deficits in crowding are going untreated, simply due to a lack of understanding regarding its effects. By increasing our understanding of the mechanisms of crowding, and its neural correlates in the amblyopic visual system, we will not only develop hallmarks for screening the presence of crowding but also neural markers for examining the effect of treatment programs on the reduction of crowding and its effects on the brain. This will reach not only clinicians, including orthoptists and ophthalmologists, but also the patients themselves.

Our work will also impact upon treatment programs for amblyopia. Firstly, the demonstration of 'amblyopic zones' in the normal periphery, with large interocular differences (IODs) in crowding, will allow the potential to develop 'treatment programs' to reduce these IODs. These treatment programs will mirror the 'binocular' treatments currently in development for amblyopia, which focus specifically on the training of stereo-vision (depth perception) and/or the reduction of interocular suppression. This is in contrast with current 'patching' techniques that are exclusively monocular and focus only on reducing the IOD in acuity. Because interocular suppression and impaired stereo-acuity are also believed to correlate with elevated crowding, we will specifically seek to evaluate the effectiveness of our analogous treatments in reducing the IODs in crowding (as well as acuity). This will allow the development and evaluation of these programs without the need to involve the central vision of children, which is time consuming and risky.

The development of these training programs for normal vision will then feed directly into 'binocular' treatment programs currently being developed within Moorfields Eye Hospital. I am a collaborator on this project, with Prof. Steven Dakin, which is funded by the Special Trustees of Moorfields and a UCL Impact Award. These exploratory trials are aimed at developing a binocular treatment that reduces interocular suppression through daily training. Our evaluations in the normal visual field in this project will determine which skills are likely to be most effective at being trained to reduce crowding, which will dictate the direction of the clinical trials. The impact of our findings will thus be immediate.

The project will also impact upon the screening of amblyopia. One of our aims is the further development of the Vac-Man battery of visual tests for children. These videogame-based tests are aimed at examining children's acuity, contrast sensitivity, stereo-acuity and, importantly for the project at hand, visual crowding. The use of these tests in the current project will allow their further refinement and validation as tools for clinical use. There is a great lack of age-appropriate tests for complex visual functions in children and this project would certainly fill the gap.

Further, an exciting aspect of these screening tests is their similarity of these tests to video games. Through contacts that UCL has established with game designers in Electronic Arts, I will attempt to develop these tests towards videogames designed to be distributed over the internet as tests for the screening and monitoring of crowding. The aim will be an industry collaboration that will be developed towards the end of the project.