One of the most exquisite tasks performed by the visual system is the precise discrimination of depth. Because of its finesse, it is reasonable to assume that stereopsis may deteriorate at very early stages of progressive visual system damage. Hence, stereo testing is worth considering as a component of a critical test battery.
Stereopsis is subdivided into two categories, local and global. The local category depends exclusively on central vision and local category tests depend on small test fields and small viewing angles, i e, minute targets and/or long viewing distances. Global tests have much the opposite properties as exemplified by the many varietes of random dot and vectographic stereograms. Global tests involve more widely distributed central processing and global stereopsis can fail without affecting local stereopsis [1, 2]. Defective stereopsis may be restricted to certain portions of visual space as illustrated elsewhere on this site [3, 4].
As to local stereopsis, normal observers are commonly held capable of discriminating an image disparity of 10" (seconds of arc) in central vision, corresponding to a discrimination in depth of about 0.15 mm at 0.3 m viewing distance and 1.5 mm at 1 m. Many subjects perform even better, up to five-fold better. Hence, the eternal problem of wide normal limits crops up again. Clinically practicable local stereopsis tests are hard to come by. Several computer-based tests have been developed in research settings [e g, 5]; none seems to be easily available. Of course, the classical approach of using movable rods or falling beads is still viable. The little calculator presented below may be helpful when considering optimum parameters. Great care must be taken to avoid monocular clues.
Global stereopsis tests are commercially available in many variants. The main applications relate to strabismus and amblyopia and little is known about the utility in neuro-ophthalmological conditions. Most tests employ printed plates and separate the right and left eye images by anaglyph or polarization techniques. Normal limits differ between the various tests and range over some 20" to 80" [6].
The screen-based random dot "V" test presented here depends on inexpensive anaglyph filters (red for the left eye, cyan for the right), as used to view anaglyph photographs and films. Old-time red-green goggles will also work.
The left-hand panel demonstrates the test task, namely, to identify the orientation of the arrowhead-shaped target. The right-hand panel shows the test display. Here, crossed disparities generate a reference background just in front of the screen surface and an arrowhead target that appears to protrude from the background. The target orientation is varied at random and points up, down, right or left. The orientations are numbered clockwise from top (1 - 4): the current orientation is indicated by a digit below right in the display, together with the current disparity (0 - 6 units).
Use the above figure to explain the test task before actually opening the test. When the test has been opened, left-click on the test panel to decrease the disparity, right-click to increase the disparity. Slide the cursor outside the panel to generate a new target, without changing the disparity.
Disparity can be adjusted in units of 2 picture elements. Most LCD screens use a pixel size of about 0.29 mm. Hence, the smallest disparity is about 2 x 0.29 mm, corresponding to an angular separation of 60" at 2 m viewing distance.
A characteristic feature of random dot stereograms is that the perception of depth rarely is instantaneous. Indeed, several seconds may be needed for the depth percept to emerge, particularly with disparities close by the threshold. It may be helpful to turn the gaze to something else for a short while. The time to perception may hold useful diagnostic information. This has been explored in several scientific fields but apparently not within neuro-ophthalmology.
Interestingly, stereo vision can be assessed objectively by means of dynamic random dot stereograms, using pupil reactions as an indicator [7].
A general overview of the topic of color vision is presented elsewhere on this site. Here, optimum conditions for identifying minimal degrees of acquired dyschromatopsia in central vision will be discussed.
Again taking the neural matrix model as the starting point for the presention, the question is how to best detect a minimal depletion of neural channels carrying color information. These channels are not spread uniformly across the matrix. Lack of uniformity is most striking for the relatively rare blue channels, which make up some 10% of the color channels. The blue channels are completely missing from the foveal part of the matrix. Overall, red channels are somewhat more common than green channels, making up about 50% vs 40% of the total population. Hence, if any given extra-foveal color channel could be tested in isolation, there would be a 50% probability that it would be found tuned to red, 40% to green, and only 10% to blue. As usual, there appears to exist considerable variation between individuals.
One way to quantitatively assess a depletion of color channels could be to use a resolution-related test like visual acuity, using color targets. Unfortunately, the rendition and calibration of color is extremely complicated, both in print and by other means, and explorations largely have been restricted to laboratory settings. Not unexpectedly, studies of normal subjects using the most common types of laboratory target, i e, gratings, have revealed substantially lower acuities for color gratings than for achromatic ones, and particularly low acuities for blue gratings
[8].
Studies of subjects with acquired dyschromatopsias presently seem to be lacking.
What may argue against the use of gratings and similar, spatially extended targets, is that they will envelop large numbers of channels, causing an over-abundant stimulation and conceivably a low sensitivity to damage. Possibilities of targeting small numbers of channels, or even individual channels, should be better with point-wise sampling, e g, with narrow laser beams and adaptive optics [9, 10] but clinically practicable instruments are not availabe as yet. The simpler approach of using color micro-dots in the fovea test included in rarebit perimetry has met with little success because of limitations of current liquid crystal displays.
It is hard to find good arguments to support the common use of tests like Ishihara's pseudo-isochromatic charts and Farnsworth's D-15 in critical color testing. Neither test was devised for acquired dyschromatopsia. On the contrary, Ishihara's test was expressly devised to detect congenital defects and the D-15 to assess vocational aptitudes. Neither test was devised for quantitative assessments.
Currently, the best approach to critical diagnosis appears to be to capitalize on the fact that acquired dyschromatopsia very rarely is perfectly symmetrical between the eyes and that the patient is the best judge of the presence or absence of asymmetry. Any asymmetry should be easiest to detect against a perfectly neutral background, using colors from those parts of the color circle that are known to be most severely affected by neurological disease. The use of fuzzy target borders should help to focus attention on the color aspect and divert attention from border definition, which is a distracting variable. Flashing should help to minimize after-images and Troxler fade-from-view effects. The Color Comparison Display attempts to optimize test conditions according to these principles. Test distance and refractive correction are not critical. The single target serves comparisons between the eyes, in conjunction with alternating occlusion. For within-eye comparisons, multiple targets are more useful.
Complaints of discomfort from brilliant lights are well-known symptoms of optical faults and diseases of the macula and need no comment here. However, at the other end of the spectrum, in the case of unexplained visual loss, actual testing for a subtly increased sensitivity to glare and/or a subtly prolonged glare recovery time may be helpful.
The Glare Test is presented on a separate page to minimize distracting features. The test should be run in a completely dark space. Seat the subject at a reading distance, with reading glasses, if applicable. Maximize the browser window to minimize potentially disturbing lights.
Start the test by identifying the minimum brightness required to correctly identify the slowly alternating test letters shown in the display's center. The letters are drawn at random from the HOTV series. For the examiner's convenience, the current letter is shown also in the upper left corner of the display. Letter brightness can be decreased by tapping the keyboard's down-arrow key and increased by tapping the up-arrow key. It is also possible to change the letter brightness by left- or right-clicking the mouse inside the circular test area. The current brightness level is shown to the right of the letter indicator (top left). Once the threshold brightness has been identified, make a note of the result. From this point on, there are two ways to proceed:
Because of the vagaries of individual computer displays and the difficulties of exact calibration, normative values have to be defined individually for each set-up, from test results obtained from normal subjects. The luminance values shown in the display are nominal only and do not necessarily belong to a linear scale. The test letter size corresponds to approximately 0.2 decimal (Snellen 20/100) at 300 mm viewing distance.
| Page 1 2 3 |