The greatest development in the field of low vision assessment in decades was a direct result of observing an international competition for blind cricketers.

In October 2008, one of the world’s leading pioneers in the field of vision testing and research was invited to a meeting in England to discuss the vision standards being applied by the World Blind Cricket Council (WBCC). Dr. Ian L. Bailey accepted the invitation, which opened the door to this breakthrough in low vision testing.

The Berkeley Rudimentary Vision Test (BRVT) – About its Development and its Applications. 
By Ian Bailey, Emeritus Professor of Optometry and Vision Science; University of California, Berkeley

I began my career as a low-vision clinician in 1972, and right from the start, I rejected the common clinical practice of counting fingers (CF) and using hand motion (HM) as the way to test visual acuity when the patient’s vision is too poor to be measured with the usual acuity charts. The testing procedures and the visual tasks involved in the CF and HM tests were neither well-defined nor well-controlled. In the early days, my preference was to bring a Bailey Lovie visual-acuity chart, or sometimes a Feinbloom chart, very close to the patient when I could obtain a visual acuity score based on the combination of the viewing distance and the size of the letters that were read. It is somewhat embarrassing for me to acknowledge that it was not until 2007 that I gave serious attention to the measurement of very poor visual acuity.

My attention to measuring very low vision was triggered by my friend and colleague, Hasan Minto, a Pakistani optometrist when he invited me to join an Advisory Committee on Vision Standards for the World Blind Cricket Council. We met in Brighton (UK) in October of 2007. Visual acuity was used as the WBCC’s primary vision criteria for classifying their visually impaired cricketers. Our committee was asked to review their WBCC classification criteria and the various testing procedures that were being used. One of the classification criteria then being used by the WBCC and most other blind sports organizations was “unable to recognize a human hand at any distance”. It was immediately obvious that there was some need for change.

Two weeks later, I was one of about 40 invitees to attend a meeting that Art Jampolsky organized at the Smith-Kettlewell Eye Research Institute (SKERI) in San Francisco. The meeting was called “Visual Prostheses: Toward Standards for Outcomes and Their Evaluation.” Most of the exploratory research being done with early visual prostheses used CF and HM results as the primary outcome measure. Michael Bach, a German vision scientist, had already developed the Freiburg Acuity and Contrast Test (FrACT) (1996, OVS, 73, 49-53), and he made it universally available as a free download.  

The FrACT is a battery of computerized vision tests that offers several different optotypes, variable contrast, and tests of more basic vision functions, and it has a sound computerized psychophysical methodology for determining thresholds.

A couple of years previously, I had been involved in testing children at Schools for the Blind in Malawi, and I was acutely aware of a need for vision tests that were very simple, easily portable, did not require access to electricity or batteries, and could be used by almost anyone in almost any environmental conditions. Immediately after the SKERI meeting concluded, I put together a PowerPoint program that used Single Tumbling E’s (STE’s) and square wave gratings as visual acuity targets. To test more basic vision functions, I included some targets with large white areas against black backgrounds. Our testing protocol had the display screen at presented two viewing distances – one was equal to the height of the screen, and the other was 4 times the screen height. There was
an initial set-up procedure to avoid shape and size distortions. A week later, this new PowerPoint routine was being used in the UC Berkeley low vision clinic, but we were well aware that we needed something simpler and more efficient. The PowerPoint file contained 40 Single Tumbling E’s (5 different STEs at each of 8 different sizes), and there were 12 square wave gratings ( 4 each at 3 different spatial frequencies). There were also 12 targets for the more basic visual localization
tests.

At the longer viewing distance, the PowerPoint STE test gave a visual acuity range of logMAR 2.00 to 1.30 (20/2000 to 20/400) in increments of 0.1 log unit. With the closer viewing distance, the range became logMAR = 2.60 to 1.90 (20/8000 to 20/1600). The grating acuity test was only performed at the closer viewing distance, and there were only 3 sizes logMAR = 2.90, 2.60, and 2.30 (20/16000, 20/8000, 20/4000). We liked this PowerPoint test. It gave nice results, but it was extremely slow and clumsy. We wanted a visual acuity test with the targets printed on cards and did not require electrical power or batteries.

At this time, Jonathan Jackson from Belfast was visiting Berkeley on sabbatical leave. Over the next year, Jonathan and I spent countless hours cutting and pasting, making multitudes of square and rectangular cardboard panels with various optotypes and patterns printed in different sizes. We experimented with many different combinations and sequences. In the Low Vision Clinic, Robert Greer, Marlena Chu, and I experimented with different combinations of these cards and panels. Hasan Minto was in Pakistan, and we regularly consulted with him. Gradually, our group reached an agreement on some fundamental principles; we made our choices of targets and sizes. This new test is called the Berkeley Rudimentary Vision Test (BRVT). 

The basic principles were:

1. There is a limitation of the range of visual acuity that can be measured with a letter chart.
2. Beyond the limit of the letter chart, the visual task must be simplified. Reading single optotypes is a simpler task than reading optotypes in chart format. Identifying the orientation of gratings is simpler task than identifying single optotypes.
3. Very close viewing distances would be required when it became necessary to have the test targets at very large angular sizes.

Limitations of the acuity range with letter charts?
There is a limit for measuring visual acuity on charts of letters or other optotypes. The logMAR chart design from Bailey and Lovie (1976) standardized the visual acuity task using 5 letters per row, a logarithmic size progression, and proportional spacing. There should be minimal variation of legibility within the family of optotypes. The patient’s task on such a chart is to read across the rows of letters.

The ETDRS chart (1982) uses the logMAR principles and has a 4 meter testing distance with the Sloan letter family as the optotypes. The ETDRS chart has become the universal “gold-standard” for serious measurement of visual acuity. At 4 meters the top row of the ETDRS chart has letter that are art the level of logMAR = 1.0 (20/200 or 6/60), and the top row subtends an angle of about 7.5°. When the patient is unable to read across the top row, the chart should be moved to a closer distance so that the angular size of the letters becomes larger. We consider that 1 meter is the closest distance at which the ETDRS chart should be presented. At 1 meter, the top row subtends an angle of about 30° degrees (±15°). Any wider, and observers are likely to employ head turns and
body twists to read from one side of the chart to the other. Also, at 1 meter, there is a 3.5% difference between the distance from the eye to the edge of the chart and to the center, and this is marginally worrisome. At one meter, the top row with its 40M letters has a logMAR value of 1.60 (20/800, 6/240).

A technical but significant point is that the top row does not have any letters above it, so it presents a slightly different task. Its letters are less embedded or cluttered. So, strictly speaking, the limit of letter chart visual acuity that can be measured with an ETDRS chart at 1 meter is reached when the patient is unable to read any letters of the second top row. That is, the visual acuity should be better than logMAR = 1.60 (better than 20/800 or 6/240). It is often recommended that when the chart is placed at 1 meter, a +0.75D lens be used to maintain the best optical focus. It is extremely unlikely that such fine differences in focus would have any impact on the ability to recognize very large targets (20/800, 6/240)). When the patient cannot read the top of a logMAR chart at 100 cm, the visual task should then be made simpler.

SIMPLIFYING THE VISUAL ACUITY TASK

Visual Acuity with Isolated Single Optotypes 
Using single optotypes was the natural choice for a visual task that is simpler than reading a chart of optotypes in the logMAR format.

While Sloan letters are the most favored optotypes for logMAR charts, it would be cumbersome to use isolated Sloan letters. This would require 10 different letters at each size level. For Landolt Rings and Tumbling E’s, only one optotype is required at each size level, and they can be rotated to give different orientations (usually 4) that the patient is required to identify. The Landolt Rings and the Tumbling E’s are familiar and well-accepted optotypes. We consider that the Tumbling E is the better choice of the two because it is a more complex target. With the Landolt Ring, when its size is close to the threshold, the observer does not actually see the gap but rather perceives a small blob or spot that is not quite symmetrical. From the perceived asymmetry, the observer may infer the location of the gap in the ring. There is more complexity in the Tumbling E, which is constructed with three parallel limbs and a crosspiece. At the threshold, the Tumbling E target appears as a small square-ish blob, and the observer’s task is to identify any patterns, irregularities, or features that help identify the orientation of the Tumbling E target. One can never tell what features are being used. How much of a clue comes from the parallel limbs? Does the crosspiece make the blob darker on one side? Do the ends of the 3 limbs create some irregularity on one side? The complexity of the Tumbling E task seems more like the complexity of reading letters when the size is close to the threshold. For example, if the observer is trying to make a choice between two relatively similar letters (say H and N), what irregularities or asymmetries within the blob provide critical clues to the identity of the letter from which the blob was derived?

We chose the Single Tumbling E’s as the single optotype. We decided that the largest target should be at a visual acuity level of logMAR = 2.60 (20/8000, 6/2400). This angular size is achieved with a 100M letter E (14.5 cm high) viewed when from 25 cm. Then, the E subtends about 30°. When the patient cannot read the largest STE at 25 cm, the visual task needs to be made simpler.

Visual Acuity with Gratings
Square wave gratings were the obvious choice for a simpler test target. Gratings have a long tradition in visual acuity measurement. Like optotypes, the fundamental structural elements are closely spaced with black and white lines. But, unlike optotypes, the visual task of recognizing the orientation of the lines of the grating, the observer does not need to restrict attention to a specific region or to analytically construct a coherent shape. Control of fixation is more or less irrelevant. The observer’s visual task is simply to recognize that somewhere within the display field, there are some stripes with a particular orientation. Are there any stripes or lines anywhere? Are they horizontal or vertical? It does not matter whether the stripes being detected are close to the observer’s point of visual attention or whether they are seen with peripheral vision. It is easy to understand how patients with substantial central vision loss often have great trouble identifying letters (or other optotypes), but they can have relatively little difficulty in recognizing the presence and orientation of a grating somewhere within a large display. The coarsest grating included in the BRVT has two black and two white stripes, and each stripe is 6 cm wide (subtends 14°at 25 cm). One black-white cycle is 28°. The finer gratings have proportionately more stripes.

When the coarsest of the gratings is presented at 25 cm, it has a visual acuity value of logMAR = 2.90 (20/16000, 6/4800 ). If the patient is unable to identify this grating, the measurement of visual acuity is terminated. Testing continues with tests of more basic vision functions. 

Basic Vision Function Tests.
One test is for white field projection (WFP). There are two large display fields (52° square). One display panel is black with a white quadrant; the other panel is half black and half white. The observer’s task is to identify the location of the large white area. Both cards should be presented in 4 different orientations. 

The other test is for Black-White Discrimination (BWD). Again, there are two large 25 cm square cards at a viewing distance of 25 cm. One card is all black, and the other is all white. The observers’ task is to tell whether the card is black or white. 

SOME PRACTICAL CONSIDERATIONS

Card Size
If the visual acuity targets are going to be on printed cards, how large should the cards be?

We chose to make all the test cards the same size – 25 cm (10 inches) square. Having the test cards all the same size makes handling them a little easier, and the 25 cm dimension makes them easily portable. Mainly for mathematical convenience, we had decided that 25 cm should be the closest recommended viewing distance, and the 25cm cards could become a convenient measuring tool for checking this short eye-to-chart distance.

Sizes of the Visual Acuity Targets
There were competing pressures. We wanted to cover a large range of possible visual acuity scores; we wanted the size increments to be reasonably small, but we did not want to have many cards.

Single Tumbling E’s
We chose to make the Single Tumbling E test with two cards hinged together, so there are only 4 STE targets, one on each side. The chosen sizes were 100M, 125M, 800M and 500M. In millimeters, the height of these four E’s were 15cm, 9.5cm, 6 cm and 3.8 cm.

The recommended testing routine uses only two viewing distances, 100cm, and 25 cm.

At the 1-meter testing distance, the 4 E targets have visual acuity values of (logMAR = 2.00, 1.80, 1.60 and 1.40, or in Snellen format 1.0/100, 1.0/63, 1.0/40. 1.0/25

In “Equivalent Snellen” terms, these are 20/2000, 20/1250, 20/800, 20/500 or 6/600, 6/380, 6/240, 6/150.

At the 25 cm meter testing distance, the 4 E targets have visual acuity values of logMAR = 2.60, 2.40, 2.20, and 2.00, or in Snellen format 0.25/100, 0.25/63, 0.25/40. 0.25/25

In “Equivalent Snellen” terms, these are 20/8000, 20/5000, 20/3200, 20/2000, or 6/2400, 6/1500, 6/950, 6/600.

Thus, the range of STE acuity becomes logMAR = 2.60 to 1.40 (20/8000 to 20/500, or 6/2400 to 6/150). The size increments are in steps 0.20 log units (size ratio=1.60/1). These size increments are not as fine as those used on logMAR letter charts where the steps of size are 0.10 log units (ratio 1.25/1 = 5/4)

PRODUCT PROFILE

Assessment and Rehabilitation are critical to those who suffer from Low Vision and can help them learn how to stay independent and make the most of their sight. A diagnosis and eventual rehab can help your patients with low vision live full, active lives.

Browse through our highly standardized vision tests which cover the vast array of Low Vision assessment tools found exclusively at Precision Vision.

ASK PRECISION VISION

Q: With Sloan being such a popular optotype, why did you choose Tumbling E for the BRVT?

A: While Sloan letters are the most favored optotypes for logMAR charts, it would be cumbersome to use isolated Sloan letters. This would require 10 different letters at each size level. For Landolt Rings and Tumbling E’s only one optotype is required at each size level and they can be rotated to give different orientations (usually 4) that the patient is required to identify. The Landolt Rings and the Tumbling E’s are familiar and well-accepted optotypes. We consider that the Tumbling E is the better choice of the two because it is a more complex target. 

You may read more here.