Colour tests have been with us for a long time - but new techniques promise a revolution in detecting early serious disease.

About 1 in 8 males has an inherited defect of colour vision. While this is important to a boy who wants to be a police marksman, or a seaman, in most situations the defect does not matter very much and no-one tests routinely for such problems. Older people, with severe blinding eye diseases, also suffer from colour vision defects.- but again we tend never to test for them. But there are very good theoretical reasons why colour vision tests should be of use, and as a result of developments in computer games, it is now possible to carry out these tests simply, effectively, quickly and economically.

The basis of seeing in colour is of course the three different types of cones in our eyes. But when the cone pigments were first analysed, there was a great surprise. The three pigments are not equally spaced over he spectrum. The “ blue” pigment (short wave length) is a very good blue, but the two remaining pigments both absorb best in the green part of the spectrum: there is no long wavelength pigment which allows us to see red. Although the quality of redness seems so unique, the sensation is due to layer after layer of neural processing, that begins in the retina and continues into the higher visual brain centres. It is not surprising that anything that interferes with the computation should affect colour vision. Furthermore, 90% or more of the fibres of the optic nerve are concerned with colour processing. They belong to the ‘X system’ If we only test vision with black and white (as with an optotype ) most of the time we are testing the Y system and exclude most of the visual neurones! The black and white system is of great importance: it is essential for peceiving motion, for detecting shading , for stereopsis and depth perception, and can handle all detail up to a surprisingly high level, equivalent to a visual acuity of > 0.5. The “private pathways” by which the foveal cones send coded colour information can also handle ultra-fine detail – but are not often needed. In fact, the only tasks the Y system cannot handle are to help thread a needle and detect when fruit is ripe. But the quality of our visual sensation is entirely dependent on the colour coded X system. The greatest obstacle to using colour tests in the clinic is the lack of good, rapid, sensitive tests for blue yellow colours, because then we can eliminate all the congenital disease and concentrate on acquired blinding conditions. Fortunately blue vision suffers earliest in almost all retinal disease, possibly because the blue cones are very fragile, but also because they are very few in number. A single blue cone has a large receptive field. If it, or its ganglion cell dies, there is a significant “hole” for blue in the visual field, and proper tests of colur vision can demonstrate this easily. But until recently all tests of blue vision were relatively non-quantitative, and slow. The situation has been revolutionised by using computers to present coloured images to the patient, using the techniques developed for use in computer games.

Colour- the basics
The test in practice looks deceptively simple: a coloured letter (optotype) is flashed on the screen. The letter is visible only because it has a different hue to the background on which it appears. The difference in hue can be made more or less obvious, by moving the hue along a predetermined line in colour space. One of these lines is the red –green axis that is important for congenital colour deficiency and the line is the protanopic colour confusion line. To persons who lack one of the red-green cone types, all colours on this line are confusible. Another colour confusion line, at right angles to the first, is a blue-yellow or tritan colour confusion line. Persons who completely lack blue cones confuse all colours on this line. The two lines form the minor and major diameters of an ellipse, known as the MacAdam ellipse, after the colour scientist who first made the measurements. All the colours inside a MacAdam ellipse seem the same to a normal observer. In eye disease, one of the axes increases in length. In congenital eye disease, the minor axis increases in length, and in acquired retinal disease it is common to find a large and disproportionate elongation of the major axis. It is simple for a computer to alter the distance along a colour confusion line between the background and optotype, and thus alter the visibility. The minimum distance at which the letter can be recognised gives the colour threshold. Modern psychophysical methods mean that thresholds can be determined very quickly- in about the time it takes to measure visual acuity one can find the protan and tritan thresholds.
Take the test

When one investigates more deeply, there are tricks that are employed to improve the clinical value of such a test. There has to be close control over the brightness and colour properties of the system, or else, it might be possible for a patient to use any artefactual brightness to make the distinction. Since we all have different amounts of yellow pigment in our lenses and in the macula, such variations are bound to occur, but they can be masked by having an ever-changing pattern of brightness on the screen , to remove any brightness clues. The pattern must flash briefly, for then, the patient cannot make eye movements to scan the letter. If the size of the optotype is chosen correctly, small defects- perhaps the loss of a single blue cone- will make it impossible to recognise the letter of the optotype, even if the colour is clearly visible. When the optotype is made very large, of course more than just the fovea is tested, and this is important, because diseases like diabetic retinopathy need not begin in the foveal retina. In other forms of the same test, the optotype is discarded, and the patient is presented with a fixation point, while targets are confined to the retinal periphery. The arcs of the retina where glaucomatous field losses occur than thus be selectively probed . Normally when an image is presented to the retinal periphery it quickly fades with time ( the Troxler effect). With the flashed images produced by a computer this does not happen.

To take such a test is simplicity itself. The operator chooses the type of disease to be investigated, and the computer does the rest. It prompts the operator for patient details, and show the letters to the patient. Only approximate refractive correction is required. The patient names the letters (or an illiterate E or symbol can be used), and the operator notes whether the response is correct. The computer homes down on the threshold, and whn it is satisfied, terminates the test. It displays the treshold, and relates it to normal age-corrected values. The print –out gives a coloured indication: blue if the threshold is normal, green, yellow, orange or red for increasing losses.

What results can be expected? In the 3 major blinding disease ( Glaucoma, Diabetes and Age-related maculopathy ) detection of the condition can be made very early. Blue scotomas develop before black-and white ones. The test is much less noisy and easier for the patient, because instead of determining thresholds at many locations, one can concentrate on just a few. The results are as good as those obtained by methods that examine the optic disc . In diabetes, colour vision begins to be lost before there is any visible change in the fundus. By the time that even a few micoraneuysms are visible, the thresholds are significantly elevated. To conduct such a rapid screening test is far preferable to dilating the pupils, waiting, and conducting a detailed search with an ophthalmoscope. The test is more sensitive than any other, including fundus photography with sophisticated digital systems. Furthermore, progression can be monitored; by the time a background retinopathy has deteriorated to a sight threatening condition, the titan threshold will be much elevated, and protan thresholds will begin to rise. at this stage, of course, specialist intervention is required; but up to this point, a lay person can administer and interpret the test results.
The same is true for age-related maculopathy. Even in very sophisticated societies, it is difficult to detect or monitor the progress of ethe early stage of this condition. However, blue losses begin when only a few drusen, or a slight irregularity of pigmentation can be seen. By the time the patient is in danger of an acute degenerative change, a flashed optotype of a size that more than covers the fovea is indistinguishable no matter how intense the colour. Now that new treatments are proposed for ARM, the problem of early detection and selection is becoming important. Nor is this all; a vriety of conditions, from drug intoxications, inherited degenerations and optic neuropathies also produce extra-ordinary losses of colour vision. For example, Stargadt’s disease alone of the retinal degnerative conditions frequently spares blue cones. Dysthyroid eye disease is best monitored by losses of blue-yellow discrimination, and the success of surgical or medical intervention can be seen within 24 hours. Optic neuropathies classically affect all colour systems equally, so if there is a disproportionate loss of blue, then retinal damage has also occurred. In summary, there are inexhaustible reasons for colour vision testing, and its possible to screen evryone who requires a new prescription. Optometrists have reason to be grateful for the computer-games industry!

The base system consists of an industry standard Workstation Personal Computer (PC) running the Windows operating system.

A dual output graphics card is installed to provide dual screen display, (operator consol and patient stimuli).
The ChromaTest program is installed on the computer's hard disk.
Hard copy is provided by a colour ink jet printer, whilst patients data and results are stored in an SQL database.
Calibration is carried out using a simple photometer on the stimulus display.
Additional equipment consists of a CCTV camera for monitoring patients fixation.

A guide price for the basic system would be in the region of £5800.00.