There are many good reasons to consider periodic self-examinations of vision. For example, there is a clear advantage in early detection of insidiously progressing and potentially blinding conditions like age-related macula degeneration (AMD) and glaucoma. These widely spread conditions are notoriously difficult to detect by self-observation. Unfortunately, early detection by formal testing faces a severe obstacle in the form of an immense variation in test results between normal subjects. Exactly where in this normal range any given subject normally belongs cannot be ascertained. Therefore, so-called normal limits have limited relevance for detection of low-grade abnormality in the individual case. One solution to this dilemma is to perform periodic examinations and to look for ominous trends among test results. Arranging for frequent expert examinations is not realistic.
It might be thought that the ubiquitous visual acuity test would be an ideal tool for self-detecting and self-monitoring both AMD and glaucoma. Unfortunately, this is not the case. Acuity tests are well suited for detecting and correcting optical faults, and for checking for compliance with various types of licensing requirements, but they are not well suited for detecting and grading retinal or neural conditions. A new test paradigm may prove more effective but awaits clinical validation [A]. Visual acuity will be left aside in the following.
There exists a multitude of apps and websites that attempt to replicate standard clinical tests. For reasons explained under the Technical Background heading, this is not necessarily the best approach to self-testing. The tests presented here take new routes. They are based on detailed deliberations presented elsewhere on this site [B,
C, D].
It is not necessary to dig into their technical backgrounds to use the tests presented here. However, it is important to recognize that no test, be it clinical or home-based, can be 100% correct. Each and every test can occasionally produce false-positive or false-negative results. A professional opinion should always be secured whenever test outcomes indicate a problem or appear questionable. Once a professional diagnosis has been secured, self-testing can be resumed to monitor the evolution of the disorder over time.
The tests presented here come in two versions: a quick one for pass-fail testing, and another, more time-consuming one for quantitative assessments. The latter version is capable of storing the test result. Most tests contain control mechanisms for assessing the user's power of observation. They need a display size of at least 15 x 15 cm (6" x 6"). Make sure that the browser's magnification ("zoom") is set to 100% by pressing CTRL and 0 (zero) keys.
Although the tests are presented under AMD and Glaucoma headings, they are equally applicable to other disorders affecting the macula and the visual pathways. Begin by selecting the category of interest:
Most clinical tests of vision depend on rigidly standardized instruments that ensure constant test conditions. Standard conditions are very difficult to realize with display-based replicas because different displays vary with respect to resolution, luminance, contrast, gamma, color rendition, and other properties. Calibration is very difficult to arrange outside laboratory settings. Even such a basic aspect as viewing distance is hard to control. For optimum utility, display-based vision tests had better explore non-traditional avenues, where exacting calibrations can be deferred. Obviously, test results obtained under such unconventional circumstances may be difficult to compare with those produced by standard clinical tests. Essentially deriving their roots from the 19th century, the latter tests may actually not be the best [B,
D].
A thorny issue associated with conventional clinical tests is an enormous variation among normal subjects. Take visual acuity as an example. 1.0 (or 20/20, or 6/6) is generally held to define the normal level but most normal subjects actually perform much better. Indeed, most everyone with an acuity result of only 1.0 has the right to complain [E]. Notably, there is also variation within one and the same subject on repeated examinations.
Yet another problematic aspect of conventional tests is that they generally have a poor sensitivity to low-degree neural damage. This is likely attributable to an overload of information in the test targets.
One way to potentially improve test performance is to capitalize on the fact that the visual system is very adept at recognizing deviations from symmetry. Test targets can be constructed to exploit symmetry and the lack thereof. One approach is to arrange minuscule target components in regular patterns and ask for signalling of any perceived asymmetry. Minuscule target components, or "rarebits" [F], also have the strong advantage of minimizing information overload.
The new tests presented here aim to minimize both information overload and normal variation, without the need for exacting calibrations. The tests contain control mechanism that illuminate the test subject's power of observation. The tests are thought to be fairly robust against variations between different types of display devices.
Macular disorders commonly generate subnormal visual acuity, microscotomas, and metamorphopsias (distortions of shape and size). Visual acuity is well known to correlate poorly with the severity of the underlying disorder and will be left aside here. Mapping of microscotomas requires dedicated hardware so self-testing is limited to qualitative observations (see further). Assessment of metamorphopsia is the only practicable alternative for quantitative self-testing. There are three variants of metamorphopsia: micropsia, macropsia, and mixed micro/macropsia. Dysmetropsia is an alternative name for metamorphopsia.
The conventional approach to metamorphopsia assessment is to examine a print of a regular grid with a central fixation mark. The task is to trace any apparent irregularities with a pencil, while looking at the fixation mark all the time. This so-called Amsler test is widely spread in spite of decades of severe criticism in the scientific literature (see [1] for a recent report). The present approach involves two radically different routes.
The pass-fail route introduces a moving hexagonal test pattern, which continuously sweeps the test area and dynamically highlights any deformations. The second, quantitative route replaces the clutter of a grid with a symmetrical geometrical shape, which in case of apparent deformation can be restored with the aid of easily manipulated computer controls. Additionally, and uniquely, any deviations of size can be measured by comparison with a reference shape placed far outside the macula.
1. Pass-fail test
The MetaScreen display presented in the left panel below uses flowing hexagons rather than a stationary grid. The symmetrical flow is expected to enhance fixation stability at the same time as it seamlessly sweeps the test area and counteracts the normal Troxler fade-from-view effect. Click on the display to start movement, click again to stop. Except for the frame rate there are no adjustable parameters. The test task is to identify any deviations from regular hexagonal shapes, while fixating on the red center dot. The right panel shows a simulated snapshot of what a localized dysmetropia might look like. A MetaScreen variant using flowing circles can be viewed
here. Interestingly, central fixation is not necessarily the best way to detect localized dysmetropia: try changing and holding fixation on a point outside the center.
MetaScreen also presents a "microscotoma" in the form of a small blank spot in a randomly selected location. Note that the blank spot is much easier to detect when the display is running than when it is static. Restart to make the blank spot appear in another location. If no blank spot is seen, more training is needed. This is a unique control feature that illuminates the user's power of observation.
Use a viewing distance of 0.4 m. Cover one eye. Use reading glasses, if applicable. Fixate on the red dot. Try different flow rates.
Interpretation Shape and size deviations are primarily associated with macula edema. Very rarely, macula edema is perfectly symmetrical and will not clearly affect shape perception. Safeguarding with the quantitative MetaDial test should be considered. Seeing more than one blank spot indicates the presence of true microscotomas. Any microscotomas and/or deviations from circular shape should be viewed as strong recommendations for professional follow-up.
The MetaScreen display induces a powerful counter movement upon stopping the flow. This is an example of the so-called waterfall illusion [2].
The MetaDial test replaces the clutter of a test grid with a symmetrical geometrical shape, as shown in these schematic figures:
Legend:
A - Amsler-type grid as seen by a normal eye
B - ditto as seen by an eye with metamorphopsia
C - ditto as seen with a solid square added to the grid, hiding the central grid lines
D - ditto without a grid. Inset: procedure for nulling of deformations
There are two test tasks. First, any apparent shape deformations should be corrected by clicking on the control buttons. Second, the size of the test target should be compared to that of a second shape placed far below the macula, at 15° eccentricity. In case of disparity, the size of the latter shape should be adjusted so as to obtain a perfect size match.
The test opens with recommended settings, which include the use of a hexagonal rather than quadratic test target. There are some optional settings, including other target shapes and simulations of shape deformations. Play around with the controls: nothing can get damaged.
Use a viewing distance of 0.4 m. Cover one eye. Use reading glasses, if applicable.
Following completion of the test, the results should be saved for future reference. The results include the adjusted graphic shape, the sum of unsigned adjustments, and the test target's relative size. Normal eyes are expected to show a sum close to zero and a size close to 100%.
The test result presented to the right was produced by a subject with macular scarring caused by vitreous detachment and shows what adjustments were needed to null his macropsia variant of metamorphopsia: 4 corner positions were adjusted for an unsigned sum of 0.6° and size was adjusted by 135%. Inset: arrows show directions of shape adjustments.
The MetaDial test assesses metamorphopsia along a single target border. A fuller picture can be obtained by repeating the test with different target sizes. The number of target corners can be set to the desired fineness of detail.
Interpretation Metamorphopsia is most often due to macula edema. However, there are other causes. The actual cause can not be diagnosed by self-examination but requires professional evaluation. Very rarely, macula conditions are perfectly symmetrical and will not clearly affect shape perception. Therefore, do not skip size matching.
Insidously progressive visual field defects (VFD) are a hallmark of all members of the glaucoma family of disorders. The VFDs come in many shapes and depths but tend to favor the field areas above and below the macula, forming so-called arcuate scotomas. Other disorders of the visual pathways may generate similar or dissimilar defects
[G]. The general approach to detection is one and the same.
Differential diagnosis of VFDs requires both detailed topographic maps and ancillary professional examinations.
Because ancillary examinations cannot be realized in self-test settings, there is little point in spending much time and effort to try to obtain detailed maps. It is arguably better to aim for a reasonable overview and a tolerable examination time. This can be achieved by using rarebits, that is, minuscule white dots that are briefly presented on a black background. Rarebits have proved very useful in several clinical applications, including Rarebit Perimetry [H].
The tests presented here capitalize on some novel observations:
The visual system is very adept at recognizing deviations from symmetry
Rarebits are easily arranged in symmetrical patterns
If one or more of the component rarebits cannot be seen, symmetry will be lost
Symmetrical arrangement around the fixation point aids stable fixation
The pass-fail test and the quantitative test use slightly different patterns, as detailed under their respective headings. Both tests depend on full-screen formats. Their tested areas extend to 10° of eccentricity, thereby avoiding interference by the blindspot.
Use a viewing distance of 0.4 m. A shorter distance may lead to interference by the normal blindspot (which is situated to the right for a right eye, and vice versa). Cover one eye. Use reading glasses, if applicable. Make sure that both the glasses and the display are perfectly clean: smudges may invalidate the test.
Note that all visual field tests require concentration and a bit of training. They are all critically dependent on a stable fixation. It is good practice to disregard results from the very first trial. In case of uncertainty of outcome, compare impressions between the two eyes, which normally are identical, or repeat the test at a later date. There is no point in cheating.
1. Pass-fail test
The 4RB pass-fail test employs multiple rarebit-defined squares of fixed size. If one or more of the four rarebits cannot be seen, it is easy to recognize that symmetry has been lost. Such a loss signals the presence of a VFD, and the more severe the VFD, the more rarebits will be missed. And the more squares affected, the more extensive the VFD.
A key feature of the pass-fail test is the flashing of the test targets. Flashing counteracts the so-called Troxler effect, a physiological phenomenon that causes static targets located outside the line of sight to fade from view. Exposure time is 1 s.
In order to avoid distracting lights, the 4RB test should be used in full-screen mode. This is realized by clicking the button in the upper right screen corner. There is a second button controlling rarebit size. This should be set to the smallest possible. Larger sizes may reduce the test's sensitivity to damage.
The test presents a static central fixation mark and six perpetually alternating groups of four squares, for a total of 96 - 1 non-overlapping rarebit presentations. The "- 1" term signifies that one randomly selected position is rendered blank and serves as a check on the observer's power of observation. If the resulting asymmetry escapes detection, more concentration is needed. Restart the test to obtain a new control location.
While steadily fixating on the crosshair, observe each group of flashing squares, without looking directly at any one square, and take note of any asymmetry of appearance within each group. It is not necessary to attend to individual squares because any depleted squares will directly generate an asymmetric appearance of the group as a whole. Let the 4RB test loop as many times as needed to make a decision on perfect symmetry or lack there-of. There is no time limit. The outcome of the test is a mental picture, which can be memorized but cannot be saved.
Interpretation Normally, all test targets will be seen as regular squares, except for the control target. The presence of two or more clipped squares points to the existence of a VFD. Such a result should be viewed as a strong recommendation for professional follow-up. A simple sketch of the approximate positions of irregularities may be helpful in the consultation. Turn to the quantitative test if a permanent record is desired.
The 3RB quantitative test employs 24 rarebit-defined like-sided triangles of fixed size that all point towards the fixation point. Following each of the 24 target presentations, the number of rarebits seen needs to be recorded on the inbuilt response pad. The test will not proceed until a response has been entered. Hence, the examination time is not fixed. One specific advantage of this approach is the elimination of the all-the-time fixation that is required in conventional field tests. All-the-time fixation is not only unnatural but also tiresome and boring.
It is legitimate to ask why the 3RB test employs triangular targets rather than quadratic ones, as used in the 4RB test. The reason is that the present test involves actual counting, which is a bit more demanding than judging the Gestalt of groups of targets.
Target exposure time is 200 ms. The total number of rarebit presentations equals 72 - 1. The "- 1" term signifies that one randomly selected position is rendered blank and serves as a check on the observer's power of observation. If this control presentation escapes detection, more concentration is needed. Restart the test to generate a new control location.
In order to avoid distracting lights, the 3RB test should be used in full-screen mode. This is realized by clicking the button in the upper right screen corner. There is a second button controlling rarebit size. This should be set to the smallest possible. Larger target sizes may reduce the test's sensitivity to low-degree damage. The very smallest rarebit size is obtained by changing color from white to green. White pixels light up three "subpixels" (red, green, and blue) that are positioned side-by-side. Using a single subpixel effectively reduces target width by two-thirds.
Interpretation Normally, all test targets will be seen as regular triangles, except for the control target. Areas of defective vision are characterized by inability to see some rarebits. The fewer the number of rarebits seen, the deeper the VFD, and the larger the number of defective targets, the more widespead the VFD. The spatial distribution of the defect may be classified with the help of a pattern library [G].
In the graphic summary of results each test target is represented by three closely spaced circles. Open circles represent rarebits seen and closed circles represent rarebits not seen. Normally, all, or nearly all, circles should be open. There is one exception, marked by an asterisk: this is a control presentation, containing only two rarebits.
The 3RB test result shown below was obtained from the left eye of a subject with a relative lower arcuate VFD. Note the remarkably short test duration.
Like the original RareBit Perimetry test [H], the 3RB test calculates a Hit Rate statistic. This is defined as the number or rarebits seen divided by the number of rarebits shown. Normal eyes generally produce Hit Rates in the 90 - 100% range. Due to normal age-related changes in the visual system, somewhat lower rates are expected in the elderly.
It is not necessary to dig into their technical backgrounds to use the tests presented here. However, it is important to recognize that no test, be it clinical or home-based, can be 100% correct. Each and every test can occasionally produce false-positive or false-negative results. A professional opinion should always be secured whenever test outcomes indicate a problem or appear questionable. Once a professional diagnosis has been secured, self-testing can be resumed to monitor the evolution of the disorder over time.
A thorny issue associated with conventional clinical tests is an enormous variation among normal subjects. Take visual acuity as an example. 1.0 (or 20/20, or 6/6) is generally held to define the normal level but most normal subjects actually perform much better. Indeed, most everyone with an acuity result of only 1.0 has the right to complain [
The conventional approach to metamorphopsia assessment is to examine a print of a regular grid with a central fixation mark. The task is to trace any apparent irregularities with a pencil, while looking at the fixation mark all the time. This so-called Amsler test is widely spread in spite of decades of severe criticism in the scientific literature (see [
Following completion of the test, the results should be saved for future reference. The results include the adjusted graphic shape, the sum of unsigned adjustments, and the test target's relative size. Normal eyes are expected to show a sum close to zero and a size close to 100%.
The 4RB pass-fail test employs multiple rarebit-defined squares of fixed size. If one or more of the four rarebits cannot be seen, it is easy to recognize that symmetry has been lost. Such a loss signals the presence of a VFD, and the more severe the VFD, the more rarebits will be missed. And the more squares affected, the more extensive the VFD.
The 3RB quantitative test employs 24 rarebit-defined like-sided triangles of fixed size that all point towards the fixation point. Following each of the 24 target presentations, the number of rarebits seen needs to be recorded on the inbuilt response pad. The test will not proceed until a response has been entered. Hence, the examination time is not fixed. One specific advantage of this approach is the elimination of the all-the-time fixation that is required in conventional field tests. All-the-time fixation is not only unnatural but also tiresome and boring.
It is legitimate to ask why the 3RB test employs triangular targets rather than quadratic ones, as used in the 4RB test. The reason is that the present test involves actual counting, which is a bit more demanding than judging the Gestalt of groups of targets.
