GB2355540A

GB2355540A - Visual acuity chart display and measurement apparatus

Info

Publication number: GB2355540A
Application number: GB9924273A
Authority: GB
Inventors: Ian E Murdoch; D Alistair H Laidlaw; Dan A Rosser; Fred W Fitzke
Original assignee: University College London; Moorfields Eye Hospital NHS Trust
Current assignee: University College London; Moorfields Eye Hospital NHS Trust
Priority date: 1999-10-13
Filing date: 1999-10-13
Publication date: 2001-04-25
Also published as: GB9924273D0

Abstract

A visual acuity display apparatus 14 displays at least a portion 16 of a chart comprising several lines, each with two to four optotypes. The ratio of sizes of two consecutive lines is constant. Another provided apparatus displays a chart with a crowding bar 7. A provided acuity measurement apparatus 15,18 receives, records and recognises responses given by a subject 13. A further apparatus terminates a test according to the subject's responses. Another apparatus scrolls lines of optotypes according to the responses. Another apparatus replaces only an optotype in a certain position (figs. 4a,4b). Another display apparatus selects optotypes at random.

Description

2355540 AUTOMATED VISUAL ACUITY TEST APPARATUS

FIELD OF THE INVENTION

The present invention relates to eye-charts used for measuring the visual acuity of a subject as may be done in a research setting or in a clinical setting for example in a hospital or before the purchase of spectacles or contact lenses.

BACKGROUND TO THE INVENTION The Snellen acuity chart is the most widely used form of chart used in clinical practice. In an acuity test the subject is asked to read the lines of letters, each line of which is progressively smaller down the chart, to find the smallest that can be read. This chart has inherent design flaws which result in imprecise and poorly repeatable measurements. Among these flaws are ones related to "crowding" - the environment of neighbours surrounding a letter - varying legibility of the particular letters employed and the non-uniform scale of letter sizes used (as opposed to the uniform logarithmic scale appropriate to acuity measurement). These flaws and the non-uniform Snellen fraction measurement scale render acuity data difficult to analyse and the test insensitive to changes over time in a subject's acuity.

The form of visual acuity chart preferred for research is the ETDRS logMAR chart. The ETDRS chart is named after the "Early Treatment Diabetic Retinopathy Study" for which it was originally produced. This chart has fourteen lines of five letters each. In this chart the logarithm to base 10 of the size of the letters in each line reduces by 0.1 (which is a reduction in size by a factor of approximately 1.259). This chart addresses the problems of the Snellen chart stated 2 above, the ETDRS chart providing precise repeatable measurements and readily analysable decimalised acuity data. This ETDRS chart is, however, considered widely too cumbersome for routine clinical use and has not been adopted for such use. Another chart similar to the ETDRS chart is the Dailey Lovie chart.

The present inventors have previously proposed a "Reduced logMAR" chart which greatly reduces the time needed to take an acuity test compared to both the ETDRS chart and the widelv used Snellen chart, hut which provides a much more repeatable and precise measurement than the Snellen chart and is only slightly less precise and repeatable than the ETDRS chart. A patent application has been made in the United States of America in respect of the "Reduced logMAR" chart; it is described herein as it is the preferred form of chart for use in the present invention.

Computer systems which display eye-charts on a monitor are known.

SUMMARY OF THE INVENTION According to a first aspect of the present invention there is provided a visual acuity chart display apparatus comprising means for generating a display of a visual acuity cnart or different parts thereof from time to time, and wherein the visual acuity chart comprises a plurality of lines of optotypes, wherein the ratio of the size of the optotypes of each line of the plurality to the size of the optotypes of preceding line is a constant, and each line of the plurality has between two and four optotypes.

3 Displaying a chart of this type in such an apparatus results in an acuity test being particularly rapid. Further, having a small number of optotypes on each line means that the optotypes can be large which minimises the problems of 5 pixelisation inherent in many forms of display.

Preferably a line of optotypes displayed has three optotypes but may have two or f our.

Preferably at least three lines of the chart are displayed at any one time.

The said ratio may be such that the difference between the logarithm to base 10 of the size of the optotypes of each line displayed and the logarithm to base 10 of the size of those of the next in the display is 0. 1, the next being the neighbouring line in the display or the next displayed subsequently.

The said optotypes may be drawn from the Sloan set.

According to a second aspect of the present invention there is also provided a visual acuity chart display apparatus comprising 25 means for generating a display of a visual acuity chart or different parts thereof from time to time, wherein the visual acuity chart comprises a plurality of lines of optotypes and wherein the apparatus is arranged to display a crowding bar spaced from the end of one of the said lines of optotypes.

The visual acuity chart display apparatus may be arranged to display at least one crowding bar spaced above the top line of optotypes or below the bottom line of optotypes.

4 The display may have only one or two lines of optotypes but preferably has three lines. Having more than three lines is also possible. 5 The to) line above which, or the bottom line below which the crowding bar is displayed may be the line to be read by the sub-'ect. J The crowding bars displayed may be joined by linking sections. This helps to avoid the subject from misreading the crowding bars as optotypes. The liking sections may be at an angle form the vertical on the chart.

A said crowding bar is preferably spaced from the end of a row by less than, preferably by half, the width of the optoLypes of that row, but may be spaced at other distances.

A said crowding bar is preferably spaced above the top line of optotypes or below the bottom line of optotypes by less than, preferably by half, the height of the optotypes of the top or bottom row respectively, but it may be spaced at other distances.

The breadth of a said crowding bar may be equal to the width of the strokes of the optotypes.

The apparatus may also comprise means for receiving and recording responses from a subject. This further increases the rapidity of an acuity test.

According to a third aspect of the present invention there is further provided a visual acuity measurement apparatus comprising means for generating a display of a visual acuity chart or different parts thereof from time to time, and means for receiving and recording responses from a subject, wherein said means for receiving and recording responses is arranged to receive a vocal input from a subject in respect of a particular optotype in the display of the visual acuity chart, to recognise that input and to record whether the subject has correctly identified that optotype.

The use of voice recognition also results in a particularly rapid acuity test, and further reduces or eliminates the need for the intervention of an operator (i.e. someone other than the person being tested).

The apparatus may be arranged to allow the subject of an acuity test to control the progress of the test with vocal input.

The acuity measurement apparatus may be arranged to accept both a vocal input and another form of input.

The acuity measurement apparatus may be arranged to determine whether the vocal and the other input are consistent.

The acuity measurement apparatus may be arranged to accept vocal input when performing an acuity test and to calculate an acuity measurement from the vocal input making an allowance for possible errors in recognising the vocal input on the basis of a determination of whether the vocal and the other form of input are consistent.

6 The acuity measurement apparatus may be arranged to refuse voice input if said determination is not sufficiently favourable.

The acuity measurement apparatus may be arranged to learn from known vocal inputs from an individual subject or from a particular population, adjusting itself to improve its recognition of that subject's vocal inputs.

The said adjustment to improve recognition of vocal inputs may be on the basis that the vocal inputs are same as corresponding inputs made by another form of input.

According to a fourth aspect of the present invention there is also provided an acuity measurement apparatus comprising means for generating a display of a visual acuity chart or different parts thereof from time to time, wherein the apparatus is arranged when a subject is to identify the optotypes on the next line to scroll the display.

The said scrolling may be such that the next line to be identified arrives at the same vertical position on the display as the previous 'line identified. 25 According to a fifth aspect of the present invention there is further provided an acuity measurement apparatus comprising means for generating a display of a visual acuity chart or different parts thereof from time to time, and 30 means for receiving and recording the responses of a subject, wherein the visual acuity chart has a plurality of lines of optotypes and the apparatus is arranged to terminate a test of visual acuity based on the responses of a subject.

7 The visual acuity measurement apparatus may be arranged to terminate a test of visual acuity at the occurrence of the subject failing to identify correctly a particular proportion 5 of the optotypes on a line.

The acuity measurement apparatus may be arranged to terminate the test when the subject fails to identify correctly all of the optotypes on a line. 10 The acuity measurement apparatus may derive a value for visual acuity based on the point at which the test was terminated.

The visual acuity value may be derived taking into account the proportion of optotypes correctly identified on at least one line for which the subject correctly identified at least one optotype.

The acuity measurement apparatus may be arranged to test the visual acuity of a subject a plurality of times in succession terminating each test of visual acuity based on the responses of a subject.

The number of repetitions of the test may be fixed before the test commences. Alternatively the number of repetitions of the test may determined during testing by the responses of the subject.

The acuity measurement apparatus may be arranged to start at least one of the tests after the first of the plurality at a line that is of optotypes smal ler than those of the line at which the first was started and that is larger than the point 8 at which at one of the previous tests of the plurality was terminated.

The acuity measurement apparatus may be arranged to derive an 5 acuity value -from each test.

The acuity measurement apparatus may be arranged to calculate at least one acuity statistic from the said acuity values. That may include an average, for example the mean, and the variance or standard deviation.

The tests may be repeated until the variance or standard deviation of the visual acuity falls below a certain value.

The tests may be repeated until the variance or standard deviation of the visual acuity falls below a certain value subject to a fixed maximum number of repetitions.

The present invention also provides an acuity chart display apparatus comprising means for generating a display of a visual acuity chart or different parts thereof from time to time, wherein the opt- otypes for the chart are selected at random.

Each line of the chart may have similar legibility. This may be achieved by taking into account the known legibility of each optotype and arriving at a choice which keeps the average of the optotype legibilities for a line similar form line to line.

A number of sets of optotypes may be predefined and the apparatus may be arranged to select at random a set for each line of the chart.

9 The order of the optotypes of a particular line may be chosen by the apparatus at random.

The present invention also provides an apparatus in which two or more of the preceding aspects of the invention are combined.

According to a sixth aspect of the invention there is provided a visual acuity measurement apparatus comprising means for generating a display of at least three lines of at least three optotypes means for receiving and recording responses from a subject, wherein the apparatus is arranged to replace with another optotype one of the said optotypes at a particular location after each of a plurality of responses, the location replaced being the same each time and being one that has neighbours from the plurality of optotypes to both sides on its line and both above and below.

The means for generating a display may be arranged to display exactly three lines of optotypes.

Each line of said optotypes may have exactly three optotypes.

The means for generating a display may be arranged to change at least one of the optotypes other than that at said location.

The means for generating a display may be arranged to leave the optotypes other than that at said location unchanged.

The means for generating a display may be arranged to display all of said optotypes at the same size.

The means for generating a display may arranged to display a plurality of lines of optotypes with each of said lines having a different size.

The means for generatinq a display may be arranged to scroll the display 1-o reveal a new 'line of optotypes once a particular number of replacements have been made.

The -visual acuity measurement apparatus may comprise means for evaluating a response of a subject against zhe oD.-otype at the said location.

The apparatus of any aspect of the invention may comprise a computer.

The apparatus of any aspect of the invention may have a projective display.

The apparatus of any aspect of the invention may comprise a mirror arranged to allow the subject to see the display therein, and the apparatus may be arranged to display the mirror image of each optotype.

BRIEF DESCRIPTION OF THE DRAWINGS

There will now be described by way of example only the preferred embodiment of the invention with reference to the accompanying drawings, of which:

FIGURE 1 is a "Reduced logMAR" visual acuity chart which may be used in the present invention.

FIGURE 2 is a visual acuity chart similar to that of Figure 1 but having thirteen rows of optotypes.

11 FIGURE 3 is a view of an embodiment of the present invention in use.

FIGURE 4a and FIGURE 4b illustrate an alternative method of 5 displaying an eye test chart.

DESCRIPTION OF THE PREFERRED EMBODIMENT

Figure 1 shows a visual acuity chart known herein as a "Reduced logMAR" chart for reasons which will become apparent. For ease of illustration Figure 1 has only three rows 2 of optotypes 3, for example, capital letters. In the preferred form of the Reduced logMAR chart when produced on a piece of paper or the like there are fourteen rows. (Figure 2 shows a chart of only thirteen rows.

Each row 2 of the chart has three optotypes 3, but alternatively two or four optotypes may be used. Each row of the chart is smaller in size, compared to the row above, by the same factor, which is advantageous for repeatable and easily analysable measurements. Preferably the logarithm to base 10 of the size of the optotypes reduces by 0.1 from line to line.

Preferably the optotypes are letters taken from those of the Sloan set. This set of letters, as is known in the art, are chosen to have the smallest differences in legibility (although the differences within this set are still not entirely negligible). The Sloan set is IZNHRVKDCOS'. As is described below voice recognition may be used for the subject's responses and the set of optotypes used may be limited to those that may be reliably recognised by the voice recognition system. Other sets of optotypes may be used, however, and may be more appropriate for some subjects, for example, the "Illiterate E" and I'Landholt C" sets. The 12 Illiterate E set is a capital E orientated in each of four directions separated by 900. Further the letters for a line are preferably chosen so that each line has about the same legibility as a whole, for example, a line where all the letters are the more difficult ones of the Sloan set is not chosen.

In the preferred form of chart the letters used are of a Gothic font and fit into a 5 by 5 grid of squares with the strokes of the letters being the width of one square.

Preferably the spacing 4 between two rows -Js half of the height of the lower row, and preferably the spacing between letters 5 is half the width of the letters.

There is a difference in legibility between optotypes that are surrounded by optotypes on all sides and those that are not, for examples those at the end of a row or those in the top and bottom rows. This phenomenon is known as "crowding".

To reduce legibility variations crowding bars 6 are employed. Any combination of bars to the left and/or right of rows and above the top row and/or below the bottom row may be used, but preferably all of those are used. It is also preferred to join the crowding bars 7 at the ends of the rows with linking sections 8 which are tapered if the crowding bars 7 differ in width. In most arrangements the linking sections will be at an angle to the vertical on the chart, (i.e. not perpendicular to the lines of optotypes). It is also preferable to join the crowding bars above the top line or below the bottom line to the crowding bars 7 with further linking sections 10. It is thought that crowding bars while not unhelpful to normal subjects may not be essential to their use of the chart. There are however conditions such a amblyopia for which crowding is a significant effect and for 13 which the crowding bars may well result in better test results.

The preferred widths and positions of the crowding bars are as follows. Crowding bars 7 to the left or right of a row are spaced by a distance 11 from the end of row of half the width of the optotypes of that row and have a width equal to the stroke width of the optotypes of that row. Crowding bars above the top line or below the bottom line are spaced by a distance 12 equal to half the height of the top row or the bottom row respectively and have a height equal the stroke width of the optotypes of the top or bottom row respectively.

The top line of optotypes preferably has a size of 1.0 logMAR. Larger optotypes may be displayed but the number of optotypes on a row of that size may have to be reduced so that the whole row fits within the width of the display. The logMAR unit is the logarithm to base 10 of a subject's minimum angle of resolution. 1.0 logMAR is equivalent to 20/200 in Snellen notation. Since logMAR is a measure of angle, the absolute size (for example in millimetres) of the optotypes depends on the distance from which it is intended to be viewed. Preferably this distance is marked on the chart. on a chart having a decrement between lines of 0.1 in the logarithm to base 10 of the size of the optotypes and having a first line of size 1.0 logMA.R the second line will have a size of 0.9 logMAR, the third 0.8 logMAR and so on.

Figure 2 is a prototype Reduced logMAR chart. This chart has 13 rows of optotypes with the first being of size 0.9 logMAR. (The chart of Figure 2 has been reduced to f it on the page, the chart that would be used in clinical practice being larger.) The preferred form of chart has 14 rows having an extra line at the top of the size 1. 0 logMAR. Figure 2 is 14 marked with the same reference numerals as Figure 1 where appropriate.

The following table gives the actual sizes for the letters of the chart of Figure 2. Letters for two charts are suggested, since using a chart with one set of letters for testing one eye of a subject and a different chart for testing the other reduces the effect caused by the subject remembering where the letters are. The letters used in the chart of Figure 2 are those given for chart #1 in TABLE 1. The lower rows of Figure 2 may be illegible at the reduced scale of reproduction and may be read from table 1. These charts are for -viewing at 4 metres with the top line being 0.9 IogMAR.

Line Letter Stroke Crowding bar Letters for Letters for width width separation chart #1 Chart #2 (M M) (MM) from row end 1 46.21 9.24 23.11 KNS CHR 2 36.71 7.34 18.35 SHV OZK 3 29.16 5.83 14.58 DZO NVO 4 23.16 4.63 11.58 CNK RHS 18.40 3.68 9.20 HOV QNV 6 14.61 2.92 7.31 CDZ DZO 7 11.61 2.32 5.80 OKN SNK 8 9.22 1.84 4.61 VHC VSH 9 7.32 1.46 3.66 ZSK OZD 5.82 1.16 2.91 RHO KNC 11 4.62 0.92 2.31 KVD VOH 12 3.67 0.73 1.84 SRH CZD 13 2.92 0.58 1.46 VNS DKR TABLE 1

From the above it is apparent that many charts may be obtained by varying the parameters mentioned. The preferred values are the following. The logarithm of the size of the optotypes reduces from line to line in even steps of 0.1. The chart has three optotypes per line. The optotypes are selected from the Sloan set. The chart has 14 lines with the top line being of the size 1.0 logMAR. As has been stated the absolute size of the optotypes is dependent on the intended viewing distance. The chart has crowding bars to all sides with the preferred spacings and widths mentioned above and having all the linking sections mentioned above. 5 Experiment has shown that a chart with these characteristics allows an acuity test to be made much faster than either by using the Snellen chart or the ETDRS logMAR chart, the median time for the preferred form being measured at 59 seconds compared to 110 seconds for the Snellen chart.

The preferred Reduced logMAR chart also produced acuity measurements that were more precise and hence more reliable than those produced by a Snellen chart. 95% confidence limits of agreement for two tests of the same subject (also known as test retest reliability) obtained were: Snellen +/0.33 logMAR, the preferred reduced LogMAR chart +/- 0.24 logMAR, ETDRS +/0.18 logMAR.

The preferred Reduced logMAR chart is therefore not only quicker to use than the Snellen chart but it also has better test repeatability than the Snellen chart. This is mainly because it is suited to scoring the patient's acuity per letter identified. In contrast the Snellen chart is scored per line identified because its irregularities make per letter scoring difficult. It is therefore an ideal replacement for the Snellen chart in general clinical practice. The preferred form has fewer letters on each line than does the ETDRS chart which remains preferable for research where higher precision is required. The chart is called the "Reduced" logMAR chart because it has fewer letters per line than the ETDRS chart.

16 It should be noted that the repeatability values given for the ETDRS and Reduced LogMAR charts are for "single letter interpolation" where credit is given for each letter correctly identified by the subject. This is the standard method for the ETDRS charlu and is the preferred method for the reduced Lo9MAR chart. The value for the Snellen chart is for the standard method used with that chart i.e. that of "line assignment" where a line of letters is deemed to have been read if at least one of the letters on a line is ident.-fied. Although the scoring techniques are different, it is still considered that the LogMAR charts have better repeatability than the Snellen chart.

The performance of Reduced logMAR charts with four and two optotypes per line and also with different size increments per line has been measured. The preferred form of chart for two and four optotypes is exactly as illustrated in Figures 1 and 2 with a decrement of 0. 1 in the logarithm of the optotype size from one line to the next and with crowding bars, except of course that each line has two optotypes in one form of the chart and four optotypes in another form of the chart.

A chart having two optotypes per line proved to give poorer repeatability than the preferred form with three per line and did not result in a significantly quicker test. A chart having four per line would result in a longer test than the preferred form. A chart having three optotypes per line and having a decrement in the logarithm of the size optotypes of 0.15 instead of the preferred of 0.1 provided poorer repeatability but took roughly the same time to complete a test. Therefore the preferred form of three optotypes per line and an a decrement of in the logarithm of the size of 17 the optotypes of 0.1 from line to line is thought to give the best compromise between repeatability and speed.

Of course, it is possible to have a chart that has some lines with more letters than others. For example, a chart could have the lines of the larger optotypes near the top having two optotypes, lines in the middle having three optotypes and lines of the smaller optotypes near the bottom having four optotypes.

The charts may, of course, be produced on paper for viewing with incident light or mounted on a light box. In the present invention it is preferred that they are produced as an image on a screen such as a computer monitor or are projected.

Figure 3 shows the preferred embodiment of the present invention in use. The subject 13 taking an acuity test observes a monitor 14 on which a standard microcomputer 15 displays a portion 16 of a visual acuity chart. Preferably the chart is a Reduced loqMAR chart as described above. In general a computer display will not be large enough to display an entire chart at once but preferably the monitor is arranged to display three rows of optotypes at once, the subject being invited to identify those of the middle row.

Displaying three rows provides consistent crowding effects (mentioned above), the middle row being crowded by the other two, and crowding bars 7 may be displayed at the ends of the rows (described above). The display of single optotypes or a single row of optotypes at any one time is within the scope of the invention, however.

As the test proceeds and the subject is to move on to the next row the display is changed so that the new row to be read is the middle row. If more than three rows were to be 18 displayed the row to be read should preferably not be the first or last on the screen, again for crowding reasons.

Preferably when a subject is to read a new row the display is scrolled up so that the new row is at the same height as the row that has just been read. This ensures that the subject can quickly identify which is the new row, which speeds up,-the test and reduces false readings occurring where the subject identifies an optotype other than that intended.

Crowding bars above the top line and below the bottom line on display at any particular time are not generally required because the subject 13 is not generally intended to read those lines. In some arrangements of display and distance between the display and the subject, it may be that only two rows or even a single row of optotypes of the larger sizes will fit on the screen. if the line to be read has no line of optotypes above and/or below then a crowding bar may be displayed above and/or below that row. (It may be seen from figures 1 and 2 that horizontal crowding bars 9 take up less vertical space than another row of optotypes.) Crowding bars may, of course, be used above or below a line that is to be read if it is decided for some other reason not to display a row of optotypes above or below that line.

Displays other than a monitor may be used for example a video projector for projecting an image of the chart onto a screen, where it may be observed by the subject.

Computer displays are usually divided into pixels and this places a lower limit on the size of optotype that can he displayed. Each optotype has, of course, to be resolved clearly on the screen but also substantial variations in the width of the strokes of the optotypes from the desired width 19 caused by pixelisation should be avoided. This means that the smallest optotype displayed should have strokes several pixels wide. Additionally, or if the strokes are only a few pixels wide, the display software should be arranged to make all strokes making up the optotypes of a particular line the same number of pixels wide.

With the pixel sizes commonly used, the smallest practical optotypes are larger than the smallest ones that may be printed on a piece of paper. That in turn means that f or the test to extend into the lowest ranges of visual acuity the display may have to be viewed from greater distances than those commonly used for printed charts. The extra distance may be obtained by arranging for the subject to view the display via a mirror. If a mirror is used the computer is arranged to display the mirror image ofthe optotypes so that the subject sees the correct image. At greater distance the same logMAR sized optotypes are bigger and so have more pixels in the width of a stroke. 4 metres is a typical distance at which a monitor may be used. With currently available monitors, however, it is expected that with normally good (as well as poor) acuity that a viewing distance of 2 meters will suffice.

In one aspect of the invention the computer 15 records the responses of the subject 13 by means of a keyboard or keypad 17, or by voice recognition using a microphone 18. The automatic recording of responses means that only minimal intervention by an operator is required, who may only be needed to explain the test and position the subject. The test could even be completely self administered.

The keyboard or keypad is also used before the test begins, usually by the operator, to select which test is to be taken and to select the parameters of that test.

A standard keyboard having all the characters of the alphabet may be used and for an alphabetic chart the user may simply press the key for the letter identified. Alternatively a keypad having a small number of keys may be used. For alphabetic charts the small keypad 17 may be usea' by prompting the user with a multiple choice question, i.e. one in which the subject is presented with a list of possibilities for the optotype to be identified. The subject responds by pressing a key corresponding to the one of those that he or she thinks is correct. The question may be given audibly generated by the computer 15. Alternatively, for illiterate subjects the "Illiterate E" or "LandholL C11 optotypes may be displayed and each key of -the key-pad may be overlaid with one of those symbols.

Another form of input which may be used is a pad or screen with large versions of the optotypes displayed on it. Selection of the optotype identified is made by means of a special pen or with a finger.

Voice recognition is particularly advantageous when used in the present invention since it is quick and reasonably reliable. This is because the set of optotypes from which the subject has to identify members is small in the context of voice recognition. The Latin alphabet, for example, has only twenty six characters and the Sloan set has only ten of those characters, whereas modern voice recognition systems will distinguish thousands of spoken words. This limited task for voice recognition also means that the system performs well with a wide variety of regional accents. The voice 21 recognition part of the system needs to be appropriately trained for speakers of different languages. Voice recognition can be affected by regional accents and so it may be advantageous to train the system for any particular accent. If highly accurate recognition is required, for example, in research or when following the progress of a disease, it may be advantageous to train the voice recognition system for an individual.

Whatever set of optotypes it is desired to use for reasons of legibility or otherwise, it may be that the voice recognition system has difficulty in recognising the name of a particular optotype or in distinguishing the names or two or more optotypes. Such optotypes may be eliminated from the set of those displayed either during design and testing of the system or during the testing of the voice recognition system during an initial phase (described below) of an eye test of a particular subject.

The system may be arranged so that the subject responds by speaking the name of the optotype or by giving an appropriate response to a multiple choice question. Again such a question may be given audibly.

The system may also be arranged to allow the subject to control the progress of the test. For example, if the subject says "again" the system will respond by allowing the subject to retry a line of optotypes and if the subject says "stop" the test is halted and the system alerts the operator. The system may be arranged to accept such commands from another source of input such as the keypad.

Voice recognition is not perfect and its accuracy varies from person to person. Therefore preferably the test begins by 22 displaying large optotypes that the subject is assumed to be able to read with the subject's spoken responses being recognised by the voice recognition system and the accuracy ra-e of the voice recognition system being determined. This accuracy rate may then be factored into the acuity measurement subsequently made. In this voice recognition testing stage the system may be arranged to accept another form of input, for example, from the keyboard or keypad, the subject responding to the large test optotypes both by speaking and using the other form of input. If the responses from the other form of input are correct this confirms that the subject can read the large test optotypes as assumed.

Whether or not a second form of input is employed it may be possible to infer from this initial phase of the test that the names of certain optotypes are not recognised accurately. As has been foreshadowed above, the system may he arranged to eliminate such optotypes from those that are used subsequently to make up the chart displayed.

The second form of input can also be used to train the voice recognition system. Large optotv es are displayed and the p subject is asked to make his or her input both vocally and using another form of input, for example the keypad. The system learns the particular sounds made by a particular subject for each of the optotypes which improves the accuracy of the voice recognition system. For this process the vocal input is assumed to be the same as that made by the other form so that if the subject makes a mistake in identifying one of the large optotypes that does not affect the adjustment of the voice recognition system. This procedure is particularly useful for research acuity measurements where greater accuracy of the final measurement is required.

If the test of the voice recognition system indicates that its accuracy is not acceptable the system is preferably arranged to conduct the acuity test using another form of input. 5 Once the form of input is determined the test can be explained to the subject by the operator but preferably the system is arranged to give a visual and/or audio presentation to do that which may include interactive examples.

However the subject makes his or her response the test is preferably a "forced choice" one, i.e. the subject must make a positive attempt to identify an optotype - a "don't know" response is not allowed. This removes variations in measurements between subjects who have the same acuity but where one subject is less confident in giving responses than another.

As a test proceeds the computer records each response and then may prompt the user for the next response, before recording that. Preferably the test is terminated when the user reaches a line on which he or she cannot identify any of the optotypes, i.e when he or she makes a whole line of errors. The computer then calculates the visual acuity determined by the test, displays it and stores it for future reference. The calculation may simply be to note the size of the last line on which the subject identified at least one optotype correctly. Preferably allowance is made, in order to improve precision, in the calculation for line on which some but not all of the optotypes were identified. For example, a subject failing the 0.2 logMAR line but having identified two of three on the 0.3 logMAR line would be given an acuity of 0.266 logMAR. Without such an allowance the minimum 24 resolution of a single test is the interval between rows (i. e. 0. 1 lo9MAR in the pref erred f orm of chart), The computer may perform the test once or may repeat the test on the same subject a number of times. Preferably to increase the rapidity of the test the size of the first line Qresented on the second and subsequent tests is set at some interval (for example 0.3 logMAR) above the acuity determined in the previous test. Some other rule may he used, for example, 0.3 logMAR above the weakest acuity measured in any of the previous tests. The final acuity calculated and displayed is the mean of the acuities obtained for each of the individual tests. A variance (or standard deviation) is also calculated from the individual acuities to give a confidence interval for the acuity measurement, which is then also displayed and recorded. The confidence interval allows it to be determined whether any change in a subject's acuity over time is statistically significant.

The number of repetitions may be fixed or may be determined by some rule calculated as the tests proceed. For example the test may terminate when the variance of the measurements has been reduced below a desired value and that rule may he subject to a predetermined minimum and/or a maximum number of repetitions. Alternatively, the number of repeats may depend on the variance reliability of subject responses either as determined by experiment previously for subjects in general, or of a particular class.

The repetition of tests is facilitated in two ways. First the sets of optotypes are not presented in the same order for each test which removes the effects of the patient remembering the optotypes from the last test. Preferably each set displayed should have roughly the same legibility. The members of the available sets may be predetermined or may be selected by the computer at random although preferably subject to the equal legibility constraint.

In an experiment using a Reduced logMAR chart testing a subject several times a test-retest variability (standard deviation) of 0.1 logMAR was achieved which compares favourably with even the ETDRS chart which for a single test has a test retest reliability of 0.18 logMAR. This method of repeating the test improves the sensitivity to change in a patient's acuity. This is because the acuity measured may be compared with the mean determined when the patient's acuity was last measured and confidence interval based a test retest variability for the patient also determined at that time, the test retest variability, and hence the confidence interval, being achieved by repeating the test being much better than that for the general population, which would otherwise have to be used.

Figures 4a. and 4b illustrate another example of how the chart may be presented. Instead of the patient reading along each line and the screen scrolling up when he or she reaches the end of each line, a constant display of eight optotypes arranged in a square with a ninth in the centre. The patient's task is to identify only the central optotype. Once he or she has done that only the centre optotype is changed. So for example, in Figure 4a, the patient would have to attempt to identify the central "KII, the display would then change, for example, to that of Figure 4b and the patient would then have to attempt to identify the central 11H". This arrangement that the patient cannot loose track of which optotype he or she is meant to identify, and that since the nearest neighbour optotypes are constant the crowding effects should be constant.

26 A variation would, however, he to display a new set of neighbours after each optotype has been identified, the task of the patient still being to identify the central optotype only. The tests would otherwise proceed as described above, with for example it being preferable for the patient to identify three optotypes at each size. Once the patient has attempted to identify the required number of optotypes at each size the whole display of nine optoypes is resized.

io Although the optotypes in Figures 4a and 4b are all of the same size, it would also be possible to make the top line larger and the bottom line smaller than the middle line, so as to appear like a traditional chart. In such an arrangement the display may be set to scroll up one line when the patient has finished a line to indicate progress. His task would remain to be to identify the central opLotype.

As will be apparent from the above many parameters of the system of the present invention may be varied. Those factors include the numbers of optotypes per line, the optotype size increment, the spacing between optotypes and lines, the use of crowding bars and their size and spacing and the form of input. While the values of these parameters that are preferred in general have been given above, it is considered unlikely that the preferred values for each subject or group of subjects form the population at large will be the same. The system has therefore been designed to he reconfigurable. This may be done by the operator selecting the parameters for the test or by the computer reading a set of values for a particular test from a file. Such files may he updated as experience is gained, for example, by sending an update floppy disc to the operator of an installed system.

27 The preferred parameters maybe determined from experimental runs performed by the system manufacturer or can be collected from installed systems in actual use. For those purposes the system is arranged to record performance statistics against the values of the parameters used for a test including the time taken, the test-retest reliability calculated and whether the operator considered the test successful. The statistics can also be used to monitor the performance of the equipment on site.

Claims

CLAIMS:

1. A visual acuity chart display apparatus comprising means for generating a display of a visual acuity chart or different parts thereof from time to time, and wherein the visual acuity chart comprises a plurality of lines of optotypes, wherein the ratio of the size of the optotypes of each line of the plurality to the size of the optotypes of preceding line is a constant, and each line of 10 the plurality has between two and four opto--ypes.

2. A visual acuity chart display apparatus as claimed in claim 1 wherein said plurality of lines compr1ses at least three lines each having between two and four optotypes. 15 3. A visual acuity chart as claimed in claim 1 or claim 2 wherein said means for generating a display is arranged to display at the same time at least three lines each having between two and four optotypes. 20 4. A visual acuity chart display apparatus as claimed in any preceding claim wherein said plurality of lines comprises at least five lines each having between two and four optotypes.

5. A visual acuity chart as claimed in any preceding claim wherein said means for generating a display is arranged to display at the same time at least five lines each having between two and four optotypes.

6. A visual acuity chart display apparatus as claimed in any preceding claim wherein said plurality of lines comprises at least seven lines each having between two and four optotypes.

29 7. A visual acuity chart as claimed in any preceding claim wherein said means for generating a display is arranged to display at the same time at least seven lines each having between two and four optotypes.

8. A visual acuity chart display apparatus as claimed in any preceding claim wherein a line of said plurality has exactly three optotypes.

9. A visual acuity chart display apparatus as claimed in any one of claims 1 to 7 wherein each line of said plurality has exactly three optotypes.

10. A visual acuity chart display apparatus as claimed in any one of claims 1 to 7 wherein a line of said plurality has exactly four optotypes.

11. A visual acuity chart display apparatus as claimed in any one of claims 1 to 7 wherein each line of said plurality has exactly four optotypes.

12. A visual acuity chart display apparatus as claimed in any one of claims 1 to 7 wherein a line of said plurality has exactly two optotypes.

13. A visual acuity chart display apparatus as claimed in any one of claims 1 to 7 wherein each line of said plurality has exactly two optotypes.

14. A visual acuity chart display apparatus as claimed in any preceding claim wherein the said ratio is such that the difference between the logarithm to base 10 of the size of the optotypes of each line displayed and the logarithm to 5 base 10 of the size of those of the nextu in the display is 0.1, the next being the neighbouring 'line in the display or the next displayed subsequently.

15. A visual acuity chart display apparatus as claimed in any preceding wherein the said optotypes are drawn from the Sloan set.

16. A visual acuity chart display apparatus comprising means for generating a display of a visual acuity chart or different parts thereof from time to time, wherein the visual acuity chart comprises a plurality of lines of optotypes and wherein the apparatus is arranged to display at least one crowding bar spaced from the end of one of the said lines of optotypes.

17. A visual acuity chart display apparatus as clamed in claim 16 arranged to display at least one crowding bar spaced above the top line of optotypes or below the bottom line of optotypes, 18. A visual acuity chart display apparatus as claimed in claim 17 wherein the means for generating a display is arranged to display a part of said visual acuity chart having only one or two lines of optotypes.

19. A visual acuity chart display apparatus as clamed in claim 18 wherein the top line above which, or the bottom line below which the crowding bar is displayed is the line to be read by the subject.

31 20. A visual acuity chart display apparatus as claimed in any one of claims 16 to 19 displaying two such crowding bars joined by linking sections. 5 21. A visual acuity chart display apparatus as claimed in claim 20 wherein said linking sections are at an angle to a line perpendicular to the lines of optotypes.

22. A visual acuity chart as claimed in any one of claims 16 to 21 wherein a said crowding bar is spaced f rom the end of a row by less than the width of the optotypes of that row.

23. A visual acuity chart display apparatus as claimed in claim 22 wherein a said crowding bar is spaced from the end of a row by half the width of the optotypes of that row.

24. A visual acuity chart display apparatus as claimed in any one of claims claim 16 to 23 wherein a said crowding bar is spaced above the top line of optotypes or below the bottom line of optotypes by less than the height of the optotypes of the top or bottom row respectively.

25. A visual acuity chart display apparatus as claimed claim 24 wherein a said crowding bar is spaced above the top line of optotypes or below the bottom line of optotypes by half the height of the optotypes of the top or bottom row respectively.

26. A visual acuity chart display apparatus as claimed in any one of claims claim 16 to 25 wherein the breadth of a said crowding bar is equal to the width of the strokes of the optotypes.

32 27. A visual acuity measurement apparatus comprising the visual acuity display apparatus of any preceding claim, and means for receiving and recording responses from a 5 subj ect.

28. A visual acuity measurement apparatus comprising means for generating a display of a visual acuity chart or different parts thereof from time to time, and means for receiving and recording responses from a subject, wherein said means for receiving and recording responses is arranged to receive a vocal input from a subject in respect of a particular optotype in the display of the visual acuity chart, to recognise that input and to record whether the subject has correctly identified that optotype.

29. An acuity measurement apparatus as claimed in claim 28 arranged to allow the subject to control the progress of the test with vocal input.

30. An acuity measurement apparatus as claimed in claim 28 or claim 29 arranged to accept both a vocal input and another form of input.

31. An acuity measurement apparatus as claimed in claim 30 arranged to determine whether the vocal and the other input are consistent.

32. An acuity measurement apparatus as claimed in claim 31 arranged to accept vocal input when performing an acuity test and to calculate an acuity measurement from the vocal input making an allowance for possible errors in recognising the vocal input on the basis of a determination of whether the 33 vocal and the other form of input are consistent or of how consistent they are.

33. An acuity measurement apparatus as claimed in claim 31 arranged to refuse voice input if said determination is not sufficiently favourable.

34. An acuity measurement apparatus as claimed in any one of claims 28 to 33 claim arranged to learn from known vocal inputs from a subject adjusting itself to improve its recognition of that subject's vocal inputs.

35. An acuity measurement apparatus according claim 34 wherein the said adjustment to improve recognition of vocal inputs is on the basis that the vocal inputs are same as corresponding inputs made by another form of input.

36. An acuity measurement apparatus comprising means for generating a display of a visual acuity chart or different parts thereof from time to time, the acuity chart having a plurality of lines of optotypes, wherein the apparatus is arranged when a subject is to identify the optotypes on the next line to scroll the display.

37. An acuity measurement apparatus as claimed in claim 36 wherein the said scrolling is such that the next line to be identified arrives at the same vertical position on the display as the previous line identified.

38. An acuity measurement apparatus comprising means for generating a display of a visual acuity chart or a part thereof, and 34 means for receiving and recording the responses of a subj ect, wherein the visual acuity chart has a plurality of lines of optotypes and the apparatus is arranged to terminate a test of visual acuity based on the responses of a subject.

39. A visual acuity measurement apparatus as claimed in claim 38 arranged to terminate a test of visual acuity at the occurrence the subject failing to identify correctly a particular proportion of the optotypes on a line.

40. An acuity measurement apparacus as claimed in claim 39 wherein the apparatus is arranged to terminate the Lest when the subject fails to identify correctly all of the optotypes on a displayed line.

41. An acuity measurement apparatus as claimed in any one of claim 38 to 40 wherein the apparatus is arranged to derive a value for visual acuity based on the point at which the test was terminated.

43. An acuity measurement apparatus as claimed in claim 41 arranged to derive visual acuity value taking into account the proportion of optotypes correctly identified on at least one line for which the subject correctly identified at least one optotype.

44. An acuity measurement apparatus as claimed in any one of claims 38 to 43, wherein the apparatus is arranged to test the visual acuity of a subject a plurality of times in succession terminating each test of visual acuity based on the responses of a subject.

45. An acuity measurement apparatus as claimed in claim 44 wherein the number of repetitions is fixed before the tests commence.

46. An acuity measurement apparatus as claimed in claim 44 or claim 45, wherein the apparatus is arranged to start at least one of the tests after the first of the plurality at a line that is of optotypes smaller than those of the line at which the first was started and that is larger than the point at which at one of the previous tests of the plurality was terminated.

47. An acuity measurement apparatus as claimed in any one of claims 44 to 46 wherein the apparatus is arranged to derive an acuity value from each test.

48. An acuity measurement apparatus as claimed in claim 47 arranged to calculate at least one acuity statistic from the said acuity values.

49. An acuity measurement apparatus as claimed in claim 48 arranged to calculate an average from the said acuity values.

50. An acuity measurement apparatus as claimed in claim 49 wherein the said average is the mean.

51. An acuity measurement apparatus as claimed in any one claims 48 to 50 arranged to calculate the variance or standard deviation of the said acuity values.

52. An acuity measurement apparatus as claimed in claim 51 wherein the tests are repeated until the variance or standard deviation of the visual acuity falls below a certain value.

36 53. An acuity measurement apparatus as claimed in claim 51 wherein the tests are repeated until the variance or standard deviation of the visual acuity falls below a certain value or the test has been repeated a fixed number o-IF times. 5 54. An acuity chart display apparatus comprising means for generating a display of a visual acuity chart or different parts thereof from time to time, wherein the optotypes for the chart are selected at random.

55. An acuity chart display apparatus as claimed in claim 54 wherein each line of the chart has similar legibility.

56. An acuity chart display apparatus as claimed in claim 54 or claim 55 wherein a number of sets of optotypes are predefined and the apparatus is arranged to select at random a set for each line of the chart.

57. An acuity measurement display apparatus as claimed in claim 56 wherein the order of the optotypes of a particular line is chosen by the apparatus at random.

37 58. A visual acuity measurement apparatus comprising means for generating a display of at least three lines of at least three optotypes means for receiving and recording responses from a subj ect, wherein the apparatus is arranged to replace with another optotype one of the said optotypes at a particular location after each of a plurality of responses, the location replaced being the same each time and being one that has neighbours from the plurality of optotypes to both sides on its line and both above and below.

59. A visual acuity measurement apparatus as claimed in claim 58 wherein means for generating a display is arranged to display exactly three lines of optotypes.

60. A visual acuity measurement apparatus as claimed in claim 58 or claim 59 wherein each line of said optotypes has exactly three optotypes.

61. A visual acuity measurement apparatus as claimed in any one of claims 58 to 60 wherein the means for generating a display is arranged to change at least one of the optotypes other than that at said location.

62. A visual acuity measurement apparatus as claimed in any one of claims 58 to 60 wherein the means for generating a display is arranged to leave the optotypes other than that at said location unchanged.

63. A visual acuity measurement apparatus as claimed in any one of claims 58 to 62 wherein the means for generating a display is arranged to display all of said optotypes at the same size.

38 64. A visual acuity measurement apparatus as claimed in any one of claims 58 to 62 wherein the means for generating a display is arranged to display a plurality of lines of optotypes, each of said lines having a different size. 5 65. A -visual acuity measurement apparatus as claimed in any one of claims 58 to 64 wherein the means for generating a display is arranged to scroll the display to reveal a new line of optotypes once a particular number of replacements have been made.

66. A visual acuity measurement apparatus as claimed in any one of claims 58 to 66 comprising means for evaluating a response of a subject against the optotype at the said location.

67. Apparatus as claimed in a first claim taken from one of the sets of claims: claims 1 to 15, claims 16 to 27, claims 28 to 35, claims 36 to 37, claims 38 to 53, claims 54 to 57, and claims 58 to 66, and as claimed in a second claim taken from a different one of those sets from which the first is taken.

68. An apparatus as claimed in any preceding claim comprising a computer.

69. An apparatus as claimed in any preceding claim having a projective display.

70. An apparatus as claimed in any preceding claim comprising a mirror arranged to allow the subject to see the display therein, wherein the apparatus is arranged to display the mirror image of each optotype.