The Frisby
Stereotest
View Shopping Cart
By LS Sasieni
The measure of the stereoscopic acuity has considerable significance in the
study of binocular vision, and has been used in the diagnosis and assessment of
visual dysfunctions other than that of stereopsis. In the subjective
appreciation of stereoscopic depth by fusion, there has to be a binocular
parallax, which can involve a vergence movement, and this fact can be used in
the diagnosis of small heterophorias and small heterotropias. Frisby (et al)
have described a study of 27 strabismic patients in whom microtropias and other
muscular anomalies were related to the stereo acuity as measured on the Titmus
test and on a random dot test, and referred to other investigations which had
been carried out earlier.
The Titmus Test
In the Titmus test, devised by Wirt in 1971, the subject wears a polarizing
spectacle to separate two superimposed polarized monocular images. The picture
for the left eye is polarized at 450 and that for the right eye at
1350, and the viewers correspondingly orientated. The test comprises
three subtest. The first is a large picture of a fly, in which parts of the
monocular pictures are separated by some 7.5mm, which, held at a viewing
distance of 30cm from the eyes, and assuming a PD of 60mm, represents a stereo
disparity of about 1016'--which, compared to the average normal stereo parity
of 20-50 seconds of arc, demonstrates only the presence of
gross stereopsis. This is intended as a demonstration of the test to the
patient, and, by its striking nature attracts the attention and interest of
children.
The next subtest consists of a series of nine squares in each of which are
four circles. In each set of four, one circle is made stereoscopic and the
subject has to detect which of the four is seen to 'stand out'. Each successive
set of circles represents an increasing degree of stereopsis, from 800 seconds
of arc to 40 seconds.
The third of the Titmus subtests is three rows of animal pictures. One animal
in each row appears to stand out, and the three rows represent stereo acuities of
400 seconds, 200 seconds and 100 seconds respectively.
The Frisby Test
The Frisby test is designed on a different basis from the Wirt test. Whereas
in the latter the stereoscopic effect is artificially produced by superimposing
two dissimilar pictures and separating them by viewing them through a polarizing
spectacle (Figure 1), the Frisby test presents targets which are actually 'in
depth' (Figure 2). The pictures are random dot patterns.
Random Dot Stereograms
Random dot stereograms were introduced about 1959, and in 1969 printed on
cards to be viewed in a stereoscope with some separating device. They have the
advantage of excluding the possibility of any extraneous clues contributing to
the sensation of depth--movement parallax, size, perspective, experience, etc.
The principle is the same as in pictorial stereograms except that nothing is
recognizable when viewed monocularly. In a small area of each of the dot
patterns, the dots are displaced with respect to the surrounding dots, so that
when viewed binocularly the separation between the small areas is different to
the separation of the remainder of the 'picture'.
The patterns do not necessarily have to be composed of dots. They can be
lines, letters, figures or shapes, but the essential feature is the random
arrangement over the area. Because of the random arrangement of the characters,
no difference is visible when viewed monocularly, but when seen binocularly the
central area appears at a different distance from the observer, and therefore
stereoscopically. The amount of displacement of the small area, relative to the
distance at which it is viewed, and the PD, is the measure of the stereo acuity.
The Frisby test differs again from the printed stereograms in that no
spectacles or other separating device need be worn by the patient. The targets
are actually 'in depth', being printed on the two sides of transparent plates of
different thicknesses. There are three plates, 1mm, 3mm and 6mm thick
respectively. Each plate has four squares of random dot patterns, one square
having the central portion printed on the opposite side of the plate from the
surrounding portion. Viewed against a featureless background, for example a
sheet of white paper, and with head and test held reasonably steady, the only
clue to depth perception is binocular parallax. In this case it is the
difference between the two levels, related to the distance from the observer,
and the PD, which gives the measure of the stereo acuity.
Range of Measurement
As there are four patterns on each plate, they can be presented in any one of
four positions and with either side facing the patient, so that the small area
may appear in front of or behind the level of the surrounding pattern. The
possibility of accurate guessing is thus practically eliminated, particularly as
each plate maybe presented repeatedly in various positions without the patient
being able to learn the correct response.
The test can be held at any of six distances, from 30cm to 80cm, the distance
being controlled by the use of a tape attached to the test and held by the
patient against the check. The tape is folded and marked in 10cm divisions,
facilitating accurate positioning. The six positions, combined with the three
thicknesses of plates, provide 18 values of stereo acuity, from 880 seconds of
arc to 20 seconds. These could be extended, of course, by increasing or
decreasing the viewing distance, but in practice these would amply cover the
range needed.
Identification of the Correct Square
Because of the great variety of presentations possible, it is necessary for
the tester to be able to know which is the stereoscopic square in order to
verify the subject's answers without the necessity for close inspection each
time. This is possible by discreetly touching the studs on each corner of the
plate, which also act as protective feet. On the stud nearest the correct square
is a small flattened area, easily felt, but invisible to the subject. This also
enables the test to be administered by persons without binocular vision, or with
stereo acuity not good enough to be able to check the answers by observation.
Child Patients
Such a test is extremely useful in the examination of child patients. It is
easy to administer and it has been shown2 that this type of test can
be successfully used as a screening test for amblyopia, large-angle squints,
microtropias, intermittent esotropia, accommodative esotropia and high (5.00D
and over) astigmatism, failure to pass the test indicating a high probability of
amblyopia, heterophoria, or anisometropia associated with the potential
development of amblyopia.
Stereo acuity
The measure of stereo acuity is in terms of the minimum discernible difference
of depth between the two levels, or the minimum discernible angle of disparity
between the displaced portions in a pair of stereograms. This angle varies with
(a) the distance between the dissimilar portions, (b) the distance at which they
are viewed, and (c) the PD of the subject.
If we consider the simplest test of stereo acuity, the three needle test, the
stereoscopic angle is given by the difference between the angle of convergence
when fixing either of the nearer needles and the angle of convergence when
fixing the more remote needle. The minimum discernible difference is the measure
of the stereo acuity. Figure 3 represents the principle of the three needle test
where:
1 is the viewing distance; d1 the difference in the levels of the objects (A
and B); RL or p the subject's PD. If the viewing distance is 300mm and the
difference in the distance between the nearer and further needles is 6mm, and
the PD is 60mm, then
RO = tan RAO and RO
=
tan RBO
OA OB
30 = 0.100 = tan 5" 42' 38"
300
30 = 0.098 = tan 5" 35' 58"
306
Angle RAL = 2 x 5" 42' 38" = 11" 25' 16"
Angle RBL = 2 x 5" 35' 58" = 11" 11' 56"
The stereoscopic angle is the
difference 13' 20"
If we now consider a pair of stereograms printed on flat card, in which the
stereoscopic effect is produced by a displacement of part of each picture in
relation to the remainder of the picture, and each picture is seen monocularly,
and the two are fused binocularly, as in a stereoscope, or if the pictures are
both printed in two colors and viewed through red and green glasses, as in an
anaglyph, or polarized and viewed through polarizing lenses, then the displaced
parts are seen in crossed diplopia, until the convergence is adjusted to fuse
them. This is the principle illustrated in Figure 3.
Again, the difference between the angle of convergence necessary to fuse the
similar parts of the pictures (the background) (RAL or RBL) and the angle of
convergence necessary to fuse the displaced portions (RNL), is the angle of
stereopsis (ARB or ALB).
This is the principle of the Titmus test. In the first ("fly")
subtest, the viewing distance is 300mm and the greatest separation of the
displaced parts of the picture (the edge of the wing) is approximately 7.5mm.
N is the position of the fused image of the points A and B.
CN =
ON
AC OR
CN =
300- CN
-
30
CN = 3.75 (300 - CN)
30
= 3.75 300 - CN
30
30CN = 3.75 (300 - CN)
33.75 CN = 1125
CN = 33.3mm
Which means that the edge of the fly's wing is seen apparently 33.3mm
nearer than its feet.
Now applying this to the principle in figure 3,
1 is 300mm; dl is 33.3mm; so OB is 333.3mm
Assuming p to be 60, then
RO = tan RAO
AO
RO = tan RAO
AO
RO = tan RBO
BO
30 = 0.1 = tan 5" 42' 38"
300
30 = 0.09 = tan 5" 8' 34"
333.3
The angles of convergence are, therefore,
RAL = 2 X 5" 42' 38" = 11" 25' 16"
and RBL = 2 X 5" 8' 34" = 10" 17' 8"
and the difference (the
stereo acuity):
1" 8' 8"
If we take 1 as OB, the OA is 266.6 and
30 = 0.1125 = tan 6' 25" and the angle
266.6
of stereo acuity is 12" 50' - 11" 25' = 1" 25'
Another method of obtaining the angle of stereoacuity is from the formula
(the angles being small):
p x d1 (in
radians) (1 radian = 57.29")
1"
where p is the PD; 1 the viewing distance; and d1 the difference in the
distances between the two planes.
60 x 33.3 = 0.0222 (radians) = 1" 16' 18"
300"
The Frisby Test in Practice
Applying this formula, the angles of stereo acuity represented by the three
plates of the Frisby test, held at the six viewing distances, are given in the
following table, which has been calculated assuming a PD of 64mm, and rounded to
the nearest five seconds of arc, but a difference in PD would not make any
significant difference. A copy of this table is provided with each test.
Table 1
|
Viewing Distance
CM IN |
Plate Thickness
6mm 3mm 1mm |
|
30
40
50
60
70
80 |
12
16
20
24
28
32 |
880
500
320
220
160
120 |
440
250
160
110
80
60 |
150
80
55
35
30
20 |
The Frisby test is extremely simple to apply and need take no more than a
couple of minutes. Being equally suitable for adults and children, it could well
be routinely included in the normal full eye examination.
References
-
Frisby et al. October 1975. BJO. 59. 10. Pp 545-552.
-
Frisby. 1975. Develo9pment Medicine and Child Neurology. 17. Pp
802-806.
STEREO ACUITY ASSESSMENT
Test the patient with various plate/distance presentations selected from the
table below to determine the lowest disparity value, which the patient can
reliably manage. It is suggested that the tester concentrate initially on the
usual reading distance of 40cm. Accurate measurements are best made using the
tape measure, held by the patient as illustrated in the figure to control
eye-to-plate distance, with the test wallet either rested on a table top or on
the tester's lap. There is no need however, to bother with the tape until the
tester has established the approximate distance at which the observer begins to
fail.
Important: observe the precautions mentioned above about holding the test
plates squarely, preventing head movements by the patient, using repeated
presentations until satisfied that the patient can or cannot make reliable
discriminations, avoiding reflections etc.
DISPARITIES (all values to nearest 5 sec arc)
|
Viewing Distance
cm (in) |
Plate Thickness
6mm 3mm 1mm |
|
30
40
50
60
70
80 |
(12)
(16)
(20)
(24)
(28)
(32) |
600
340
215
150
110
85 |
300
170
110
75
55
40 |
100
55
35
25
20
15 |
Record the lowest disparity, which the patient can reliably discriminate.
This stereo threshold is a measure of stereo acuity.
The above values are sufficiently accurate for customary clinical practice
but it should be realized that (as for other stereo tests) they are only
approximations to the exact disparities. Interpupillary distance will vary for
different patients, it is difficult to be completely sure about eye-to-test
distance (although the Frisby Stereotest does provide a tape measure to help
with this problem) and there will inevitably be slight manufacturing variations
in the tests themselves. In the case of the Frisby test, slight variations in
the thickness of the Plexiglas®TM plastic sheet used for the plates
means that any given set of plates is likely to depart in some degree from the
nominal real thicknesses of 6mm, 3mm, and 1mm, and hence there will also be
slight departures from the expected apparent thicknesses (of 4.027mm, 2.013mm,
and 0.67mm respectively - reductions caused by the 1.49 refractive index of the
plastic). Consequently, if the Frisby test plates are being used in a context
where it is meaningful to know the exact disparities being presented, then these
should be worked out from the usual disparity formula, suitably adjusted to cope
with the apparent depth reduction effect. This formula is:
Disparity = 206264.81 (p.d1)
-
(12)
where p is the interpupillary distance, 1 is the viewing distance, and d1 is
the plate thickness as measured with a micrometer.
Technical note: the stereo threshold is the smallest angle of disparity,
which can be discriminated by the patient. The associated stereo acuity is
technically the reciprocal of the stereo threshold.
SCREENING PATIENTS UNABLE TO GRASP VERBAL INSTRUCTIONS
-
Use the same general procedure as above but start by drawing the attention
of the patient to the circle-in-depth by placing one of his fingers on it.
-
Take the plate away, turn it around unobtrusively to a new random position,
and then present it afresh, again placing one of the patient's fingers on the
circle-in-depth. Repeat this procedure until you think the patient has grasped
the idea that the plate contains a target-to-be-identified.
-
Present the plate again in a new position, but this time encourage the
patient to place his finger on the circle by himself. If he can do so reliably
over a series of presentations, record STEREOPSIS PRESENT
-
Should the patient not volunteer clear pointing responses, a positive result
may still be recorded if scanning eye movements stop consistently at the
correct square on repeated testing. The tester can concentrate on the
patient's eyes and yet still know the position of the circle by discreetly
feeling the corner studs: the stud nearest the circle has a flat surface
discriminable by touch. The flat surface is on the same side of the plate as
the circle-in-depth.
-
If the observer consistently makes incorrect pointing responses despite
every effort to direct his attention to the correct square initially, record
Stereopsis
Not Demonstrated.
Note. While it is often easy to record a Stereopsis Present result even
for very young pre-verbal children by virtue of their consistently correct
pointing responses, it is not so easy to be as confident about a Stereopsis
Deficient result for such young subjects. For example, a run of incorrect
responses may be due to the patient not fully understanding what is required of
him, rather than because he lacks stereopsis. This is a problem common to all
currently available stereopsis tests but the Frisby Stereotest keeps such
"don't know" verdicts to a minimum. This is because it uses a natural
depth stimulus (and so avoids the need for often troublesome red/green or
Polaroid spectacles) and because it permits repeated presentations without the
patient learning the 'correct' response. It thus makes it a feasible proposition
to test children considerably younger than has been possible hitherto, sometimes
children even less than 1 year old.
Important: Observe the precautions mentioned above about holding the test
plates squarely, preventing head movements by the patient, avoiding reflections,
etc.
CAUTION
The corner studs help protect the plates when they are laid on table tops,
etc. It is desirable, however, to replace the plates in their envelopes in the
test wallet when they are not in use.
Frisby Stereotest
All Products In Alphabetical Order
Richmond Products Home Page View Shopping Cart