Research Lecture Advance Programme 2005/06

Vision Science and its Clinical Applications

Talks will commence at 1.15pm. All rooms with the CM prefix are in the Tait Building, City University, Northampton Square.

For enquiries please contact:

Abstracts

Vision Restoration Therapy (VRT) for the treatment of visual field deficits after brain damage

Patients with visual field deficits following stroke or traumatic brain injury can train their residual vision with VRT and thus achieve visual field expansions. These visual field enlargements are stable for over 3 years and they have been demonstrated using super-threshold or near-threshold perimetry. We have now studied whether eye movements and/or fixation are altered after VRT.

SLO measurements did not reveal evidence for eye movements before and after VRT greater than 1-2 degrees. When analyzing the perimetric topography of the visual border position before and after VRT, border shifts were typically non-uniform: while in some areas VF border shifts clearly occurred, in other areas they did not. Furthermore, when different perimetry measures are compared, the border shifts are not identical: while in some patients the border shift (ie visual field enlargements) may be pronounced when a simple task is used (such as super-threshold perimetry), no border shift may be seen when a more difficult task is used. Both the blind spot position and fixation quality measurements remained unchanged after VRT. Furthermore, measurements of the eye movements with an eye tracking system before and after VRT revealed no major alterations and patients show no preferential viewing towards the blind field.

Based on these observations it can be concluded that VRT does not affect eye movements nor fixation in any significant way. The recovered ability to detect visual stimuli in previously "blind" fields after VRT is therefore caused by genuine restoration of vision and this provides a novel approach for the rehabilitation of visual field defects.

Partial colour matching: A new method to measure unique hues

After Hering, colour is described in terms of six unique hues (white, black, red, green, yellow, and blue). Yet, there is no objective psychophysical procedure for establishing the nomenclature of unique hues. It is still unclear whether, say, purple is unique or not. A new method has been developed to ascertain the set of unique hues. The procedure was as follows. Trichromatic human observers were presented with pairs from a set of 21 Munsell chips with an instruction to evaluate whether they share any hue (partial colour match). Sets of chips that have identical partial colour match (matching classes), and largest sets of chips all of which partially match each other (chromaticity classes), were derived from the matrix of responses.

A chromaticity class is proved to consist of all chips which contain a particular hue (referred to as component hue). Thus, the number of chromaticity classes shows how many component hues an observer employed in their decision making. A matching class is proved to contain all the chips with identical component hues. We found four matching classes which contained only one component hue. Moreover, these classes consisted of just one chip. We concluded that these four Munsell chips (5Y, 10B, 5R, 10G) represented four unique hues.

While our results are in line with the previous studies of unique hues, the partial colour matching technique has an advantage that it is based on the data of type A, in terms of Brindley's classification.

Glaucoma genetics: a new approach

Objective

To describe the move towards genetic epidemiological approaches which are increasingly being used to complement classical Mendelian studies in glaucoma.

Main message

Classical Mendelian linkage methods are not sufficient to define the genetic basis of the common glaucomas. An approach based on genetic epidemiology will be necessary and will require large scale collaboration between molecular geneticists, clinicians and genetic epidemiologists. Such a network, EGS GlaucoGENE, has been formed in Europe.

Conclusions

• Genes found by classical linkage do not explain the bulk of glaucoma,
• Genetic epidemiology offers the promise of finding the genes underlying the common glaucomas,
• EGS GlaucoGENE has been set up to pursue this goal.

Human Myopia: Towards a preventative strategy

The nature of the mechanism that regulates emmetropization is still obscure, and the causes of human myopia are still only partially agreed.

Major areas of controversy are whether the retina can detect the sign of defocus so that a differential eye-growth response can be produced. Clearly the answer to this question will impact on treatment strategies.

Knowledge of the extent to which focussing errors are associated with the development of myopia is also needed, as are possible strategies for eliminating them.

These factors have impacted on the design of the Cambridge Antimyopia Trial which has completed recruitment of participants.

Visual Development and the Effect of Visual insult

Introduction

Visual pathway insult in childhood can have a powerful impact on development and organization of the brain. Retinal, ocular, neurological diseases, and disorders of higher level brain processing, may lead to vision impairment and subsequent cognitive impairment in children.

Studies

Using the electroretinogram ERG (full field and multifocal), colour vision visual evoked potentials (cVEP), contrast sensitivity using sweep VEP, motion VEPs and eye movement recording my lab is investigating the effect of Drug treatment (Vigabatrin and Hydroxychloroquin) on visual development as well as detecting neurovisual damage in diabetes before vascular disease, binocular mechanisms in infantile esotropia as well as developing visual dysfunctions in retinal disease.

Development

Essential to these studies is knowledge of development. We have developmental data for the ERG, sweep VEP contrast sensitivity and visual acuity as well as colour VEPs.

The effect of Vigabatrin treatment on visual development

Purpose: The antiepileptic drug vigabatrin results in retinal toxicity in about 30% of adult patient. Adults can be monitored with visual field testing which is not possible in young children. ERG changes occur as a result of the drug in children with infantile spasms. Our purpose was to establish which ERG variables worsened (change in ERG from first visit greater than 95% control repeatability) over time on vigabatrin also to establish the effect of variables such as age of initiation of drug treatment, cumulative dosage and sex of the child on the changing ERG. This talk will present use recent data and discuss the strongest ERG indicators of vigabatrin toxicity. I will discuss those factors that might result in worsening of ERG results. Knowledge of such factors will help define the population that might respond to the drug without the toxic side effects.

Recommendations

The Recommendations from the Hospital for Sick Children (Sick Kids), Toronto, are that retinal examination and ERG are useful in pediatric patients to monitor toxicity.

Sick Kids current protocol:

• ERGs and retinal exam at baseline and every 3 months after vgabatrin initiation for the first year and then every 6 months.
• Cone system ERGs show greatest changes with toxicity, specifically: amplitude of the cone b-wave response and flicker response.
• Amplitude of oscillatory potentials and rod-cone amplitude are not useful for monitoring possible toxicity.
• Advise referring Neurologist/Ophthalmologist of possible toxicity if ERG change is significant and sustained for at least 2 visits.

A wing guided by an eye: A sideways look at bird vision

The purpose of this talk is to introduce the bird as a model system for studies of vision. Although the avian eye and visual pathways have a surprising amount in common with humans, there are also many fundamental differences. Both the similarities and the differences present us with a range of possibilities for understanding, amongst other things, the function of the fovea, the importance of optical adaptations, the purpose of colour vision, eye development and the operation of visuomotor responses. In addition, the structural properties of the visual system can place significant constraints on various behaviours. A good example of this is the distinctive head bobbing of some birds when they walk. Our neuroscience studies of the visual response to a close object (the near response) have revealed that the circuitry is conserved across many vertebrate species, including birds. Recently we have discovered that manipulations of this circuitry can have a direct influence on the development of the eye and its refractive state. We believe that further investigations of these pathways from a comparative perspective will help identify general organisational principles that will in turn inform studies of experimental myopia. At the very least, however, you will hopefully come away with a better appreciation of the bird's eye view of the world and how studies using the bird may enhance our understanding of vision.

Does the addition of flanking contours help in the detection of amblyopia?

The crowding effect is the well-known decrease in visual acuity due to the presence of surrounding features (flanking bars or optotypes). Visual acuity measured in the presence of crowding features (such as in the Snellen chart, where most letters are surrounded by other letters) gives the best estimate of normal visual acuity. However, measurement of visual acuity in young children presents a challenge to the clinician. The traditional Snellen visual acuity charts are not appropriate for young children and are impossible to use for assessing pre-literate children. This talk will describe recent experiments we have been doing investigating contour interaction and crowding in childrenâs visual acuity charts. The aim of this research is to determine whether the addition of flanking contours can help in the detection of amblyopia.

Advantages and Disadvantages of Colour Blindness

X-linked (red-green) dichromats lack the function of either the long-wavelength-sensitive (L-) or middle-wavelength-sensitive (M-) cones. As a result, their colour vision is reduced from three to two dimensions; and they are unable to discriminate within the red-green dimension of colour space. This loss has consequences not only for the organisation of their cone photoreceptor mosaic, but also for the development of their post-receptoral cone pathways. Such changes can directly impact upon visual acuity and detection as well as colour discrimination per se. By comparing visual acuities and visual detection thresholds for cone-isolating stimuli in X-linked dichromats with those measured in normal trichromats, as well as blue-cone monochromats and heterozygotic carriers for dichromacy, we obtain insight into what advantages red-green color discrimination can bring to object detection and analysis and, more importantly, what disadvantages are caused by its loss.

Ten Commandments for Standardized Model Making

With apologies to Cal McCamy (Color Res. Appl. Vol. 10 [1985], p. 23), this note presents modelling rules and violations of them in existing color-and-lighting standards.

Before making the model

1. Decide clearly what question you intend to answer. [Counter Case: The CIE Color Rendering Index is maximal for any blackbody radiator, but we wanted an index whose value can exceed that of a blackbody spectrum. Also, not all blackbody radiators should have the same CRI.]
2. Have in mind (and describe) tests of your model that would allow you to falsify it if it were wrong. [Counter Case: any of the CIE Color Appearance Models (CIECAMs), which have nine model outputs for each color, and no protocol to test or even consistently to describe these outputs. And how can a subject differentiate a CAT02 match from a CIECAM02 match?]
3. Never recommend more than one model at a time without differentiating their application domains, lest a user view differences among the model predictions as a precision intended for each. [Counter Case: CIELAB versus CIELUV, both advanced in 1976 at the same CIE meeting.]

When making the model

1. If the ordering of mechanisms underlying a model is known, do not invert the ordering, at least without a clearly stated reason. [Counter Case: The CIECAMs perform Von Kries chromatic adaptation (supposed to be the second stage of processing) in one tristimulus basis, and then transform back to the Hunt, Pointer, Estevez (HPE) basis---the space of the cone fundamentals, where the processing is supposed to begin.]
2. Make sure your model doesn't give nonsense predictions upon crossing an arbitrary threshold. [Counter Case: In CIECAM02, the coordinates of white caused by some illuminants can cause division by zero in any of the sharpened Von Kries responses as well as in lightness J. Also, intense blue adaptation drives asymmetric matches outside the HPE triangle, where the signs of the coordinates change and the subsequent nonlinearity has infinite slope. This last pathology is a symptom of having failed to follow (4) above!]
3. Take care not to introduce artificial and ineradicable discontinuities into your model, lest it hang up an optimizer. [Counter Case: In CIE DE2000, Gaurav Sharma found three ineradicable discontinuities. In contrast, CIELABâs L* discontinuity is eradicated by increasing precision.]
4. Verify the required mathematical properties of your constructs. [Counter Case: The Judd-MacAdam-Wyszecki principal-component eigenvectors of daylight spectra aren't orthogonal to each other. All current CIE standard daylight spectra depend on these eigenvectors.]
5. Arrange your terminology so as not to mislead. [Counter Cases: In CIE94 and CMC, ? is not the delta of any coordinate H; also, in CMC(l:c), c must be unity to obtain sane results, but the user is misled into thinking c is an adjustable parameter.]

After making the model

1. When science advances, make your standard follow in a reasonable time. [Counter Case: CIE 1931 color-matching functions were corrected by Judd in 1951, whereupon a movement to change the CIE cmfs should have been underway. Here we are, 54 years later, with standard color-matching functions that are eschewed by the vision-research community. Shame on us.]
2. Be slow to tout the generality of your model. [Counter Case: CIECAMs have been touted as universally applicable, but they donât apply to a monochrome display in a dark surround---for which a white point is undefined by any CIECAM. The slack has been taken up by ATIS/CVO: "Human Factors Design Guidelines for Advanced Traveller Information Systems (ATIS) and Commercial Vehicle Operations (CVO)", FHWA-RD-98-057, Chapter 3.]

Intrinsically-photoreceptive Retinal Ganglion Cells and Primate Pupillary Responses

Human vision starts with the activation of rod photoreceptors in dim light and short, medium, and long wavelength sensitive cone photoreceptors in daylight. A parallel, non-rod, non-cone photoreceptive pathway, arising from a unique population of retinal ganglion cells, has been demonstrated in nocturnal rodents and, more recently, in primates. In primates, we have shown that these intrinsically-photosensitive, melanopsin-expressing ganglion cells are strongly activated by rods and cones. In rodents, these intrinsically-photoreceptive ganglion cells have been shown to serve the subconscious, 'non-image-forming' functions of circadian photoentrainment and pupil constriction. In primates, we have demonstrated that these cells play an important role both in light-evoked pupillary responses, and in the paradoxical pupil constriction that can occur after light cessation.

City University Home | Applied Vision Research Centre Home | Optometry Department Home