U.S. patent application number 14/244883 was filed with the patent office on 2014-08-07 for visual perimeter measurement.
This patent application is currently assigned to University of York. The applicant listed for this patent is Cambridge Enterprise Limited, University of York. Invention is credited to Louise Elizabeth Allen, Andrew Robert Jackson, Adar Pelah, Michael Edward Slater.
Application Number | 20140218691 14/244883 |
Document ID | / |
Family ID | 41202922 |
Filed Date | 2014-08-07 |
United States Patent
Application |
20140218691 |
Kind Code |
A1 |
Pelah; Adar ; et
al. |
August 7, 2014 |
Visual Perimeter Measurement
Abstract
In one example of the invention a patient display is provided on
which a fixation graphic or video is shown in substantially the
centre of the display, and then target graphics or video are shown
at different positions on the display about the centre. A test
subject is positioned in front of the patient display at a known
distance, and a camera captures an image of the test subject's
eyes, to allow gaze direction determination to be performed. The
captured images are displayed to a clinician user by a controlling
test application, the clinician making a determination based on the
images as to whether the test subject saw a target graphic or video
when it was displayed. The clinician user then makes an appropriate
input into the controlling test application, which also logs the
position at which the target image was displayed. By displaying
target graphics at multiple locations on the screen and logging the
clinician inputs, the controlling test application is able to build
up a plot of the peripheral vision field of the test subject. This
plot can then be graphically displayed to the clinician user, and
stored with the test results for future clinical use.
Inventors: |
Pelah; Adar; (York, GB)
; Allen; Louise Elizabeth; (Cambridge, GB) ;
Slater; Michael Edward; (York, GB) ; Jackson; Andrew
Robert; (York, GB) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
University of York
Cambridge Enterprise Limited |
York
Cambridge |
|
GB
GB |
|
|
Assignee: |
University of York
York
GB
Cambridge Enterprise Limited
Cambridge
GB
|
Family ID: |
41202922 |
Appl. No.: |
14/244883 |
Filed: |
April 3, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13261195 |
Feb 28, 2012 |
|
|
|
PCT/GB2010/001599 |
Aug 24, 2010 |
|
|
|
14244883 |
|
|
|
|
Current U.S.
Class: |
351/224 ;
351/246 |
Current CPC
Class: |
G06K 9/00604 20130101;
A61B 3/113 20130101; A61B 3/024 20130101; A61B 3/0041 20130101;
A61B 3/14 20130101 |
Class at
Publication: |
351/224 ;
351/246 |
International
Class: |
A61B 3/024 20060101
A61B003/024 |
Foreign Application Data
Date |
Code |
Application Number |
Aug 28, 2009 |
GB |
0915136.6 |
Claims
1. A peripheral vision measurement system, comprising: a subject
video display arranged to display test images to a test subject; a
camera arranged to capture images of the test subject; a user video
display arranged to display captured images of the test subject to
a user; and one or more processors arranged to control the subject
and user video displays, the user display being controlled so as to
be augmented with at least one indicator relating to the position
of a test image on the subject display.
2. A system according to claim 1, wherein the at least one
indicator is positioned in the user display at a corresponding
position to the position of a test image in the subject
display.
3. A system according to claim 1, wherein one or more
characteristics of the indicator are changeable so as to indicate
to the user the present state of a test image in the subject
display.
4. A system according to claim 3, wherein the one or more
characteristics include one or more selected from the group
comprising: a shape of the indicator; and a colour of the
indicator.
5. A system according to claim 3, wherein a first characteristic of
the indicator has a first value or state when a test image is about
to be displayed to the test subject at an indicated location, the
first characteristic changing to a second value or state after the
test image has been displayed at the indicated location.
6. A system according to claim 1, wherein the one or more
processors are further arranged to use determination data as to
whether the test subject has seen a test image when displayed on
the subject video display, said determination data being recorded
against the position of the test image when displayed on the
subject video display, whereby to collate measurement data
indicative of the peripheral vision field of the test subject.
7. (canceled)
8. A system according to claim 1, wherein the subject video display
is arranged to display a fixation image to the test subject, in
addition to the test images.
9. A system according to claim 8, wherein the fixation image and
test images are both video images, playback of at least part of the
video content being substantially continuous from the fixation
image to the test image.
10. (canceled)
11. A system according to claim 6, wherein if the determination
data indicates that the test subject has not seen a first test
image, an additional test image is subsequently shown to the test
subject at a related position on the subject display closer to the
test subject's line of sight.
12. A system according to claim 11, wherein the related position at
which the additional test image is displayed has substantially the
same or similar rotational angular position as the first test
image, and a smaller angular magnitude.
13. A system according to claim 11, wherein at least one second
test image at a different rotational angular position to the first
test image is displayed to the user before the additional test
image.
14. A system according to claim 6, wherein multiple test images are
shown to the test subject at different positions on the subject
display and multiple determination data received relating to the
multiple positions stored such that a set of peripheral vision test
results is created, the system further comprising one or more
storage media arranged to store the set(s) of test results, and
from which the set(s) of test results can be retrieved for later
viewing.
15-16. (canceled)
17. A system according to claim 1, and further comprising a
graphical user interface to permit the user to define a set of test
locations at which test images will be displayed to the test
subject on the subject display device.
18. A peripheral vision measurement system, comprising: a first
display device arranged to display a fixation image to a test
subject, and to intermittently display a plurality of test images;
and a processor arranged to collate data relating to whether the
test subject saw a test image at a particular location on the first
display device, whereby to measure the peripheral vision field of
the test subject; wherein the fixation image and the test images
are video images, playback of at least part of the content of the
video images being substantially continuous between the fixation
image and the test images.
19-21. (canceled)
22. A peripheral vision measurement method, comprising: displaying
test images to a test subject; capturing images of the test
subject; displaying the captured images of the test subject to a
user; wherein the captured images of the test subject are augmented
with at least one indicator relating to the position of a test
image displayed or to be displayed to the test subject.
23. A peripheral vision measurement method, comprising: displaying
a fixation image to a test subject on a display device;
intermittently displaying a plurality of test images to the test
subject; and collating data relating to whether the test subject
saw a test image at a particular location on the display device,
whereby to measure the peripheral vision field of the test subject;
wherein the fixation image and the test images are video images,
playback of at least part of the content of the video images being
substantially continuous between the fixation image and the test
images.
24-26. (canceled)
Description
TECHNICAL FIELD
[0001] Examples of the present invention relate to a visual
perimeter measurement system and method.
BACKGROUND TO EXAMPLES OF THE INVENTION
[0002] Accurate measurement of peripheral vision in a human can
lead to early diagnosis of potentially serious medical conditions.
Causes of visual field deficit include neonatal brain haemorrhage
or stroke and eye conditions such as retinitis pigmentosa and
glaucoma, additionally over 50% of childhood brain tumours present
with visual impairment.
[0003] A human eye converts light entering it into neural signals
which are then sent down the optic nerve. Thus, the left optic
nerve carries all of the visual information from the left eye, and
the right optic nerve carries all of the visual information from
the right eye. The optic nerves meet at the optic chiasm within the
brain. Within the optic chiasm nerve fibres carrying visual
information from the visual field temporal (lateral) to the
vertical mid-line, cross to the other side. The resultant right
optic tract therefore consists of nerve fibres subserving the
temporal visual field of the left eye and the nasal visual field of
the right eye and the left optic tract consists of nerve fibres
subserving the temporal visual field of the right eye and the nasal
visual field of the left eye these are called homonymous visual
hemi-fields. The optic tracts carry the left and right homonymous
hemi-field information to the occipital lobe of the brain. The left
homonymous visual hemi-field is processed by the right occipital
cortex and the right homonymous visual hemi-field is processed by
the left occipital cortex.
[0004] The pattern of visual field loss will therefore allow the
clinician to determine the position of a tumour or brain injury. An
ocular abnormality such as glaucoma or retinal detachment will
cause a visual field defect which crosses over the vertical
mid-line. A tumour on or touching the optic nerve will result in
loss of the central visual field in one eye along with reduction of
central visual acuity. Tumours involving the optic chiasm often do
not initially affect central visual acuity but interrupt the nerve
fibres subserving the temporal visual field as they cross within
the chiasm--causing a temporal hemi-field loss in both eyes
(bitemporal hemianopia). Bitemporal hemianopias do not usually
cause symptoms of visual disturbance but are important to detect in
order to diagnosis a chiasmal tumour at an early stage. Tumours or
brain injury involving the optic tract and occipital cortex will
cause a homonymous visual field defect--for example a tumour in the
right occipital lobe will cause loss of the temporal hemi-visual
field from the left eye and the nasal visual field in the right
eye. This would be termed a left homonymous hemianopia.
[0005] Accurate measurement of peripheral vision can be a first
indicator of brain tumours and, as described, the pattern of visual
field loss can localise the tumour or other pathology. Serial
measurements of the peripheral visual field defect can aid in the
monitoring of tumour growth and indicate if further therapy is
required.
[0006] Several techniques are known for measuring peripheral vision
of a human patient. Two common perimetry measurement apparatuses
are the "Goldman perimeter", and the "Humphrey field analyser".
However, both sets of apparatus require the human patient to place
their head into an enclosed compartment, and to maintain their
vision at a fixation point in the centre of the compartment,
suppressing their reflexive eye movement toward the appearing
target in the visual periphery. Such techniques are therefore
difficult to use with small children, for example, because small
children find the test extremely intimidating, and are unable to
maintain visual fixation at the central point in accordance with
the test instructions appropriately. Similarly, elderly patients
who may be suffering from degenerative brain diseases also present
the same problems.
[0007] In order to get around such problems various other
techniques have been developed to try and measure peripheral vision
in small children. FIG. 1 illustrates a diagram from Suga et al,
"Development of a Quantative Perimeter Screening System for Young
Children Based on an Immersive VR Display", Electronics and
Communications in Japan, Part II, Volume 89, No. 11, 2006. Here, a
virtual reality technique is used to construct a pseudo video
space. The immersive display device has three screens measuring 3
m.times.2.25 m on the front and the two sides, and also a 3
m.times.1 m vertical screen at the bottom. The system includes four
projectors to produce video images on the respective screens, four
computers to send the video signals to the projectors, and a
computer to provide synchronisation signals to these computers.
[0008] In order to provide a fixation point for the child subject,
a video image which attracts the attention of the subject is used.
Then, while the video image is being displayed at the fixation
point, image targets, which are simply round circles of light, are
made to appear on peripheral screens. A single camera is provided
focussed on the subject's face, and image processing is used to
determine the view line of the subject i.e. the direction of gaze
of the subject, using template matching. At the time a target is
displayed, if, as a result of the image processing it is determined
that the subject view line changes within one second of the target
presentation in a direction agreeing with the direction of the
target presentation, it is judged that the target has been
recognised.
[0009] The Suga et al system therefore presents an attempt at
producing an automated perimeter measurement system, using image
processing of the subject to try and determine the subject's
direction of gaze, and whether a target has been detected. However,
the system is extremely complicated, requiring much space, and
equipment to set up. In addition, the actual criterion by which it
is judged whether a target has in fact been seen is open to error,
as a decision is made purely based on whether the subject looks
towards the direction of the target, but not whether in fact the
target has been seen.
[0010] Other, manual, techniques are also known. FIG. 2 illustrates
the "white sphere kinetic arc" method. Using a white ball as a
fixation point, another white ball is moved into a child's visual
field by a clinician, and the child's response is monitored by a
second clinician. This method requires two clinicians, and the
movement of the second clinician may be distracting, leading to an
inaccurate test. As such, this method is far from ideal, delivering
a potentially poor result, as well as being costly to
administer.
[0011] There is thus a need for a visual perimeter measurement
system that can be used with small children, and that will provide
more reliable, accurate, and consistent results.
SUMMARY OF EXAMPLES OF THE INVENTION
[0012] In one example of the invention a patient display is
provided on which a fixation graphic or video is shown in
substantially the centre of the display, and then target graphics
or video are shown at different positions on the display about the
centre. A test subject is positioned in front of the patient
display a known distance therefrom, and a camera captures an image
of the test subject's eyes, to allow gaze direction determination
to be performed. The captured images are displayed to a clinician
user by a controlling test application, the clinician making a
determination based on the images as to whether the test subject
saw a target graphic or video when it was displayed. The clinician
user then makes an appropriate input into the controlling test
application, which also logs the position at which the target image
was displayed. By displaying target graphics at multiple locations
on the screen and logging the clinician inputs, the controlling
test application is able to build up a plot of the peripheral
vision field of the test subject. This plot can then be graphically
displayed to the clinician user, and stored with the test results
for future clinical use.
[0013] In one example of the invention the fixation graphic and the
target graphic is video content, which continues playback when
displayed as a fixation graphic, and subsequently as a target
graphic. In this way, a child subject's attention can be maintained
as the video content playback is continued irrespective as to
whether it is shown as the fixation graphic or a target graphic,
and hence the story represented by the video content is not
interrupted. The particular video content is selected by the
clinician user so as to appeal to the particular child test
subject.
[0014] In another example of the invention a visual perimetry test
is dynamically adapted during the running of the test by adding in
additional test points in dependence on whether the test subject
has seen the points. The determination as to whether the test
subject has seen a test point may be performed by a user or
automatically using image processing algorithms operating on video
imagery of the test subject. Howsoever such a determination is
made, in this example the test is dynamically adapted by adding in
an easier test point into the test at which a target graphic or
video is to be shown. The easier test point is added in such that
it is located closer towards the centre of the test subject's
vision i.e. has a smaller angular magnitude with respect to the
central line of sight of the test subject. The added test point may
then be displayed to the user immediately after been added, or may
be displayed to the user later, after other test points at
different angular positions have been shown. In the latter case the
continuity of the test is maintained, and the test subject should
not experience any long gap between seeing test images that may
cause the subject to wander their eyes around the screen.
[0015] In view of the above from one aspect an example of the
invention provides a peripheral vision measurement system,
comprising: a first video display arranged to display test images
to a test subject; a camera arranged to capture images of the test
subject; a second video display arranged to display captured images
of the test subject to a user; and one or more processors arranged
to control the first and second video displays, and having at least
one input arranged to receive feedback data from the user relating
to the displays; wherein the feedback data relates to a judgment by
the user as to whether the test subject has seen a test image when
displayed on the first video display, said feedback data being
recorded against the position of the test image when displayed on
the first video display, whereby to collate measurement data
indicative of the peripheral vision field of the test subject.
[0016] From another aspect an example of the invention also
provides a peripheral vision measurement system, comprising: a
subject video display arranged to display test images to a test
subject; a camera arranged to capture images of the test subject; a
user video display arranged to display captured images of the test
subject to a user; and one or more processors arranged to control
the subject and user video displays, the user display being
controlled so as to be augmented with at least one indicator
relating to the position of a test image on the subject
display.
[0017] In one example the at least one indicator is positioned in
the user display at a corresponding position to the position of a
test image in the subject display. The corresponding position may
take into account the direction of view of the camera to the test
subject. In particular, the corresponding position of the at least
one indicator on the user display may be translated in the plane of
the display in dependence on the displacement of the camera
position from the central gaze direction of the subject.
[0018] From a further aspect an example of the invention provides a
peripheral vision measurement system, comprising: a first display
device arranged to display a fixation image to a test subject, and
to intermittently display a plurality of test images; and a
processor arranged to collate data relating to whether the test
subject saw a test image at a particular location on the first
display device, whereby to measure the peripheral vision field of
the test subject; wherein the fixation image and the test images
are video images, playback of at least part of the content of the
video images being substantially continuous between the fixation
image and the test images.
[0019] From a yet further aspect another example of the invention
provides a peripheral vision measurement system, comprising: a
first display device arranged to display a plurality of test images
to a test subject at different positions on the display; a
processor arranged to control the first display device to display
the plurality of test images at the different positions, and to use
a determination as to whether the test subject has seen a
particular test image at a particular position; wherein the
processor is further arranged to calculate the position of an
additional test image to be shown to the user at a related position
to the particular position of the particular test image in
dependence on the determination, the related position being closer
to the test subject's line of sight than the particular position,
and to control the first display device to display the additional
test image at the related position.
[0020] Within one embodiment of the above the related position at
which the additional test image is displayed has substantially the
same or similar rotational angular position as the first test
image, and a smaller angular magnitude. In addition, within one
embodiment at least one second test image at a different rotational
angular position to the first test image is displayed to the user
before the additional test image. With such an arrangement, the
test subject does not become aware that test points are being added
in to the test, as there is no unusually long gap between the
subject perceiving test images.
[0021] Another aspect of the invention provides a peripheral vision
measurement method, comprising: displaying test images to a test
subject; capturing images of the test subject; displaying the
captured images of the test subject to a user; and receiving
feedback data from the user relating to the displays; wherein the
feedback data relates to a judgment by the user as to whether the
test subject has seen a test image when displayed on the first
video display, said feedback data being recorded against the
position of the test image when displayed on the first video
display, whereby to collate measurement data indicative of the
peripheral vision field of the test subject.
[0022] Further features, aspects, and examples of the invention
will be apparent from the appended claims.
BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Further features and advantages of the present invention
will be apparent from the following description of embodiments
thereof, presented by way of example only, and by reference to the
accompanying drawings, wherein like reference numerals refer to
like parts, and wherein:--
[0024] FIG. 1 is a diagram of a perimetry measurement system of the
prior art;
[0025] FIG. 2 is a drawing illustrating a perimetry measurement
technique of the prior art;
[0026] FIG. 3 is a diagram illustrating the components of a first
embodiment of the present invention;
[0027] FIG. 4 is a diagram illustrating a patient screen in the
first embodiment of the present invention;
[0028] FIG. 5 is a drawing illustrating a clinician view in the
first embodiment of the invention;
[0029] FIGS. 6(a) to 6(d) are a sequence of drawings illustrating
the operation of the first embodiment of the present invention;
[0030] FIG. 7 is a block diagram of the components of a second
embodiment of the present invention;
[0031] FIGS. 8(a) to 8(c) are drawings illustrating the arrangement
of a camera and patient screen in the second embodiment of the
invention;
[0032] FIG. 9 is a photograph illustrating the system set up of the
second embodiment of the invention;
[0033] FIG. 10 is a photograph illustrating the patient screen in
use in the second embodiment of the invention;
[0034] FIG. 11 is a photograph illustrating the clinician screen in
the second embodiment of the invention;
[0035] FIG. 12 is a flow diagram illustrating part of the operation
of the second embodiment of the invention;
[0036] FIG. 13 is a screen shot from the clinician screen in the
second embodiment of the invention;
[0037] FIG. 14 is a further screen shot of the clinician screen in
the second embodiment;
[0038] FIG. 15 is a first screen shot of the patient screen in the
second embodiment;
[0039] FIG. 16 is a second screen shot of the patient screen in the
second embodiment;
[0040] FIG. 17 is a further screen shot of the clinician screen in
the second embodiment;
[0041] FIG. 18 is a further screen shot of the clinician screen in
the second embodiment;
[0042] FIG. 19 is a perimetry plot obtained by the measurement
system of the second embodiment;
[0043] FIG. 20 is a flow diagram illustrating part of the operation
of the second embodiment;
[0044] FIG. 21 is a flow diagram illustrating another part of the
operation of the second embodiment;
[0045] FIG. 22 is a further screen shot of the clinician screen in
the second embodiment;
[0046] FIG. 23 is a further screen shot of the clinician screen in
the second embodiment;
[0047] FIG. 24 is a further screen shot of the clinician screen in
the second embodiment; and
[0048] FIG. 25 is a further screen shot of the clinician screen
illustrating a test template making tool used in the second
embodiment.
DESCRIPTION OF SPECIFIC EMBODIMENTS
[0049] Various examples of the invention will now be described with
respect to the accompanying figures.
[0050] A first example embodiment of the invention is shown in FIG.
3. Here, a visual perimetry measurement system 10 comprises a
patient display screen 3 in front of which a test subject, in this
case a small child 6, is positioned. A clinician display screen 5
is also provided. A controlling computer 4 controls the images
displayed on both the patient display 3, and the clinician display
screen 5. The controlling computer 4 is provided with a suitable
software program and corresponding hardware, in order to cause
displays to be produce on the patient display screen 3, and the
clinician display 5. In addition, the controlling computer 4 also
receives feedback from the clinician display screen 5, for example
through a mouse or keyboard interface to allow a clinician to
select buttons displayed on the clinician display, as will be
described. Also provided as part of the system, but not shown in
FIG. 3, is an image capture device such as a camera, which is
positioned so as to catch a frontal facial view of the subject 6,
whilst the subject 6 is looking at the patient display screen
3.
[0051] FIG. 4 illustrates the patient display in more detail. One
of the main problems with performing visual perimetry tests on
young children is that young children often cannot concentrate on a
fixation point for long periods of time. To overcome this problem,
in the present example embodiment a central fixation graphic 32 is
displayed in the centre of the patient display screen 3, the
graphic being chosen so as to be entertaining to a young child
subject. For example, the graphic may be a still graphic or an
animated graphic, or may, in alternative examples, be, for example,
video output, the content of the video being a popular children's'
television program, and preferably one already known to the test
subject.
[0052] In order to provide for testing of a child's peripheral
vision whilst the child is fixated on the central fixation point
32, a target graphic or animation 34 is also provided,
intermittently shown at various test points on the screen, as
determined by the control computer 4. The test graphic may be an
animated graphic, or may, alternatively, be a video graphic, such
as described previously in respect of the fixation point.
[0053] The fixation graphic 32 may be permanently displayed in the
middle of the screen, or, in alternative examples, may be only
intermittently displayed, for example it not being displayed when a
target graphic is being displayed. Likewise, the target graphic 34
is intermittently displayed at different locations. The camera (not
shown) captures images of the subject's face from a full frontal
perspective, and these images are provided, via the control
computer 4, to the clinician display 5, such that the clinician can
determine the direction of gaze of the child subject 6.
[0054] FIG. 5 illustrates one example of the clinician display 5.
In particular, the clinician display 5 presents a video display 50
formed from the video image captured by the camera (not shown) of
the child subject 6's face when looking at the patient screen 3.
The video image is overlaid by a measurement graticule 51, on which
are marked test points, illustrating the location of test points at
which the target graphic 34 will appear on the patient display 3.
For example, test points 52 illustrate locations at which the
target graphic will appear, or has appeared, in the past. The
colour (or, in other examples, some other characteristic, such as
the shape) of the test points 52 can alter depending on whether the
target graphic has been shown at those points, and whether the
child subject noticed the target graphic at that point, as will be
described later.
[0055] The clinician display 50 also includes a number of clinician
feedback and control buttons. In particular, "pause" and "quit"
control buttons are provided, which when selected by the clinician
for example with a keyboard or mouse cause the test to be paused or
to end. During the test, the main buttons that the clinician will
use are the "no" button 54, the "maybe" button 56, and the "yes"
button 58. To use these buttons the clinician monitors the video
image of the child in the clinician video display 50, and detects
whether the child's direction of gaze moves towards the location of
the test point at which a target graphic is presently being
displayed. As will be discussed, the location of the test points 52
are shown on the clinician display, and a colour coding system
illustrates to the clinician which test points are being shown,
which are about to be shown, and which are presently being shown.
Based on this information, the clinician makes an estimate using
the "no" button 54, "maybe" button 56, and "yes" button 58 as to
whether the child subject has noticed the target graphic.
[0056] More particularly, if it is clear that the child subject has
not noticed the target graphic at the present test point being
displayed, then the clinician selects the "no" button 54. If it is
possible that the child noticed the target graphic, but the
clinician is not sure, then the clinician selects the "maybe"
button 56. In this case, the test point is maintained within the
test list, and the target graphic is shown at the same test point
again later in the test. If, on the other hand, the clinician
determines that the child does spot the target graphic, and that
the child's line of gaze moves to the graphic, then the clinician
selects the "yes" button 58. Depending on the clinician's responses
using the "no", "maybe", and "yes" buttons, the clinician's
feedback is logged with respect to the target points, and the
control computer 4 is then able to build a graphical map of which
targets were spotted by the child and which targets were not. As
such, a perimetry map of the child's peripheral vision can be
obtained.
[0057] Moreover, the number of respective "yes", "no" or "maybe"
decisions made by the clinician may be logged so as to build up
over time statistical data from which the probability of detection
and confidence measures can be found. In particular the
probabilities of detection can depend on the number of `yes`
responses as a proportion of all responses, per target. These
probabilities can then be used to refine the test in future, to
elaborate for instance on the detectability of certain targets.
[0058] FIG. 6 illustrates the operation of the clinician's video
display 50 in more detail. In particular, assume that FIG. 6a shows
the start of a test in progress. The circles on the clinician
display show the target graphic locations that will be displayed to
the child. These circles can be colour coded so as to indicate to
the clinician whether the target graphic is presently being shown
at one of the locations, whether it has been shown at the location,
and whether it is about to be shown at the location. Additionally
colour coding can be used to further indicate, if the target
graphic has been shown at the location, whether the child saw the
target at that location, or not. Starting at FIG. 6(a), target
point 62 is colour coded such that it indicates to the clinician
that the target graphic is presently being shown at this location.
All other target points other than target point 64 are colour coded
to show that they are still to be tested. Target point 64 is colour
coded with a different colour to indicate that this is the next
target point that will be shown, after target point 62. At the
present point in the test as depicted in FIG. 6(a), therefore, the
child patient screen is showing the target graphic at the position
of target point 62, and the target graphic will then next be shown
at the position of target point 64, and all other target points are
positions at which the target graphic will be shown later in the
test.
[0059] Next assume that the clinician determines that the child's
gaze moves to the target graphic located at the target point 62. In
this case, the clinician can select the "yes" button which then
becomes highlighted, as shown by highlighted yes button 66 in FIG.
6(b). The control computer 4 logs the clinician input that the
child noticed the target graphic at the location of target point
62. The clinician display also shows that the next target point at
which the target graphic will be displayed is at location of target
point 64.
[0060] Once the positive "yes" response has been logged, the
clinician display then proceeds to that shown in FIG. 6(c). Here,
the colour of target point 62 has changed, to a colour that
indicates that this target point has already been displayed. For
example, the colour of the target point may change to blue. When a
target point is being displayed, it may be shown as green. A target
point that is to be displayed next may be shown as red, and, when
it is about to be displayed, as amber. In this case, the target
point 64, which is the next target point to be displayed, changes
to amber, to inform the clinician that the target graphic is about
to be shown at that point. The clinician can then prepare him or
herself to check whether the child subject's gaze moves to that
point, when the target graphic is shown at the point.
[0061] In FIG. 6(d) the colour of target point 64 changes to green,
to show that the target graphic is at that moment presently being
shown to the child subject 6 on the patient display 3 at the
position of target point 64. Target point 62 remains a blue colour,
to indicate that it has already been shown. The display then moves
around in a cycle back to FIG. 6(a), with another of the target
points changing to red to indicate that that point will be the next
target point to be displayed. During this time period the clinician
makes a decision as to whether the child noticed the target graphic
when displayed at target point 64, and selects the appropriate
input button. The clinician display then subsequently cycles
through these several states with the clinician inputting "yes",
"no", or "maybe" for each test point, until all of the points have
been tested.
[0062] It is important during a perimetry test that the child's
attention is brought back to the fixation graphic in the centre.
Thus, once the target graphic has been shown at a particular target
point, and the physician's input has been received as to whether
the child has noticed the target graphic, the target graphic is
removed from the screen, and the fixation graphic placed back into
the middle of the screen, to draw the child's attention back to the
middle. Thus, during the state shown in FIG. 6(c), where the target
point 64 is displayed as amber to warn the physician that the
target graphic is about to be displayed at this location, the
patient screen 3 displays the fixation graphic 32 in the centre of
the screen, so as to bring the child subject 6's attention to the
centre point of the patient display screen 3. In this way, the
child's gaze is returned to the centre of the screen, such that
when the target graphic is placed onto the screen at a new
position, a proper test of the child's peripheral vision can be
performed, for that particular target graphic position.
[0063] As mentioned, as the test proceeds, the control computer 4
logs which target points are spotted by the child, and which are
not. This then allows the computer to build a perimetry plot,
showing the extent of the child's peripheral vision, as determined
by the test. An example perimetry plot, being a plot produced by
the second embodiment to be described next, is shown in FIG. 19,
although it should be understood that the first embodiment of the
invention presently described would produce perimetry plots in a
similar format.
[0064] Using the example first embodiment, therefore, visual
perimeter testing of small children can be undertaken in a reliable
and controllable manner. By using an animated graphic or video as
the fixation point, and/or the target point, the child's attention
can be maintained on the test. Moreover, the test environment is
friendly to the child, as the child does not need to place their
head into an enclosed space, such as with the Goldman or Humphreys
test apparatus of the prior art. Likewise, the fixation point of
the child is selected so as to be of interest to the child such
that the child will naturally fixate on the fixation point. The
same is also true of the target graphic, in that the target graphic
is chosen so as to be of interest to the child, such that when the
target graphic is shown the child is motivated to look at the
graphic.
[0065] From the clinician's perspective, the first example of the
invention provides a repeatable and controllable test environment
for visual perimeter tests to be conducted. Information capture and
logging is readily performed by the control computer 4, based on
the clinician input. The augmented clinician video display, having
the view of the child's face augmented by an indication as to where
the target graphic is about to be shown on the patient screen helps
the clinician to determine whether or not the child's gaze moved to
the target graphic. In addition, precise placement of the target
graphic is possible, thus allowing information about the child's
visual field to be obtained as accurately as possible.
[0066] A second example embodiment of the invention will now be
described with respect to FIGS. 7 to 25.
[0067] FIG. 7 is a block diagram of the second example embodiment
of the invention. The second example embodiment builds upon the
ideas previously described in respect of the first example
embodiment, but adds several additional features in, as will be
described.
[0068] The second example embodiment of the invention has the same
system components as the first example embodiment described
previously. That is, a controlling computer 72 is provided, which
controls a clinician monitor 74 on which a clinician display image
is displayed, and a patient monitor 76 is also provided, on which
both fixation graphics 762, and target graphics are displayed. An
image capture device in the form of video camera 78 is positioned
substantially in the centre of the patient display screen 76, so as
to capture a full frontal view of a child subject when the child
subject is positioned in front of the patient display. Both the
clinician display 74 and the patient display 76 are controlled by a
dual monitor controller 728 in computer 72, the dual monitor
controller 728 being, for example, a graphics card capable of
controlling two monitors at the same time. An example graphics card
suitable for the task is the nVidia GE Force FX5200 Graphics Chip,
which provides two analogue outputs.
[0069] The camera 78 provides images to camera controller 726 in
computer 72. The camera controller 726 may be a universal serial
bus (USB) host, in which case the camera 78 may be a USB webcam or
the like. The computer 72 further comprises a central processing
unit 724 and associated components, to allow the computer to run
the system control software, control the display on the monitors,
and receive inputs from the clinician user. In this latter respect,
a mouse and keyboard controller 730 receive control inputs from a
mouse and/or keyboard, that are operated by the clinician user for
example to select buttons on the graphical user interface provided
by the clinician display 724.
[0070] As will be understood, the hardware elements of the computer
72, patient display 726, and clinician display 724 may be off the
shelf components, provided that they meet some additional
requirements, as described next.
[0071] In particular, with respect to the patient display 726, this
should possess a high enough resolution to provide comfortable
close viewing for the child subject, whilst being large enough
that, at the distance at which the child subject is placed from the
screen, the screen can display target graphics across the range of
viewing angles that are required to be tested. Typically, the range
of peripheral field view is 30.degree. to either side in the
horizontal plane, as well as in the vertical plane. A comfortable
viewing distance is typically around 60 cm, although for child
subjects a closer viewing distance can be tolerated. The screen
size required is a function of viewing distance from the screen,
and for a viewing distance of approximately 48 cm, a screen size of
approximately 98 cm (42 inches) can be used. A suitable screen,
therefore, is, for example, a 42 inch high definition plasma
screen, such as, for example, the "Panasonic TH-42 PHD 8". Various
other high definition plasma, LCD, or LED screens are readily
available of the required size, which could also be
substituted.
[0072] With respect to the clinician screen, the requirements are
not as high. The clinician screen may be displayed on a standard
computer monitor, such as a 19 inch monitor, or the like.
[0073] As mentioned, the camera 762 is positioned in front of the
patient display 76 and fixed to the screen, as shown in FIGS. 8(a)
and (b). Rather than being positioned at exactly the centre of the
screen, which is where the fixation target will be displayed, in
the present example the camera is fixed to the screen just below
the fixation point. This means, therefore, that when the child
subject is looking at the fixation point the subject is not quite
staring into the camera lens, but the difference in direction of
gaze of the child subject looking at the centre of the screen and
the image of such as captured by the camera is very small. FIG.
8(c) shows a view captured by the camera of a subject looking at
the fixation point, and it will be seen that it appears as though
the user is in fact looking directly at the camera. As such,
changes in direction of gaze can be readily detected by the
clinician.
[0074] In the above example the camera is placed as close to the
centre of the patient display as possible, although not quite in
the centre, as this is where the fixation image is to be displayed.
However, the patient image obtained has the patient apparently
looking at the camera sufficiently enough such it appears as if the
patient is staring into the camera.
[0075] In another example, however, the augmented view can be
further adjusted to compensate for the positioning of the camera
with respect to the test subject, and the clinician view of the
subject that is thus obtained, and in particular by altering the
position of the target indicators on the augmented view such that
they are translated on the augmented view in a direction opposite
to the actual displacement of the camera from the optimal centre
position.
[0076] For example, imagine that the camera is positioned at the
bottom of the patient display, rather than substantially in the
center. In this case, from the view point of the camera, when the
test subject is looking at the fixation image in the center of the
display it will appear as of the gaze of the subject is lifted
upwards, within the top half of the clinician display. If the gaze
of the subject were to move downwards to the bottom of the patient
display, then at best from the view point of the camera it may
appear as if the test subject is looking directly into the camera
i.e. the gaze is centered in the augmented view. In reality,
however, the subject gaze is directed downwards, with respect to
the patient display. On the other hand, if the subject looks
upwards from the fixation image towards an upper part of the
patient display, from the point of view of the camera and the
clinician display it will appear as if the subject's gaze moves
even higher towards the top of the clinician display.
[0077] The opposite effects are of course obtained if the camera is
positioned at the top of the patient display--the subject appears
to be gazing downwards. The same effects in the horizontal plane
are obtained if the camera is placed to the left or right of
patient display. The effects also compound with each other--a
camera placed at the bottom right hand corner will make it appear
as if the test subject, when looking at the central fixation image
will be gazing at the upper left part of the clinician display.
[0078] To compensate for camera positioning, therefore, the
placement of the target indicators on the augmented clinician
display can be adapted such that they appear at the position on the
clinician view where it will appear, from the point of view of the
camera looking at the test subject, that the subject will look if
he or she sees the test image at the target location in the patient
display. Generally, as noted this will involve a translation of the
position of a target indicator in the augmented view in a direction
opposite to the displacement of the camera from the ideal center
position (or, more generally, from the fixation image
position).
[0079] For example, if, as noted above, the camera is located at
the bottom of the patient display, in the middle, then the subject
central gaze (i.e. when looking at the center of the patient
display) will appear in the clinician view to be raised into the
upper half of the clinician view. Therefore, a target indicator in
the augmented display corresponding to a test target in the patient
display located in the middle horizontal plane of the patient
display should be displaced upwards in the augmented clinician
display into the upper half of the clinician display. Likewise, a
target indicator in the augmented display corresponding to a test
target in the patient display located at the bottom of the patient
display should be displaced upwards in the augmented clinician
display towards the middle of the clinician display. Similarly, a
target indicator in the augmented display corresponding to a test
target in the patient display located at the top of the patient
display should be displaced right at the top of the augmented view
in the clinician display.
[0080] It will be understood that similar displacements of the
target indicators in the augmented view in the opposite direction
from the displacement of the camera would also occur if the camera
is located at the top of the patient display, or to the left or
right of the display. Moreover, the displacement can also be
combined i.e. if the camera is located at the bottom right corner
of the display then the target indicators are displaced upwards and
to the left in the augmented display.
[0081] The amount of displacement of the target indicators required
depends on the displacement of the camera from the optimal central
position. Various linear or non-linear scaling factors can also be
applied to calculate the displacement.
[0082] In order to control the hardware components just described,
the computer 72 is provided with a computer readable storage
medium, such as a hard disk drive 732, on which is stored software
which when executed by the CPU 724 causes the computer to display
the clinician video display on the clinician display 74, and the
fixation and target images on the patient display 76, as will be
described in more detail. Additionally, the software enables the
computer 72 to receive feedback from the clinician via the mouse
and keyboard, to enable the clinician user to interact with the
graphical user interface provided by the clinician video display.
The software has several modules, and makes use of different sets
of stored data, as will be described next.
[0083] A first software module to be stored on hard disk 732 is the
clinician view module 7326. The clinician view module is
responsible for running the entire application, and causes the
clinician video display, including the clinician graphical user
interface, to be displayed on clinician display 74. Further details
of the clinician video display, and the clinician GUI, will be
given later. However, the clinician view module 7326 is responsible
for providing the main elements of the GUI, and in particular in
enabling the clinician to enter patient details, accessing patient
records including records of previous tests, so that these can be
displayed to the clinician, and providing GUI control buttons to
enable and capture clinician feedback during tests. Clinician view
module 7326 also controls the output of the video display captured
by the camera 78 during a test, and controls, together with the
test module 7340, the running of a test, for example to indicate to
the clinician where a target graphic is next going to be displayed
on the patient display.
[0084] A second software module is the child view module 7330. The
child view module 7330 interfaces with the clinician view module
7326, and acts under the control of the clinician view module 7326.
The main task of the child view module 7330 is to control the
images displayed on the patient display 76. The child view module
7330 accesses stored content, stored as themes 7322, which are
content for the fixation and target graphics. As discussed
previously, the fixation and target graphics may be animated
graphics, or may, preferably, be video content, such as children's'
television programs, or the like. This content is stored
categorised by "themes", on the hard disk 732, as the themes
records 7322. An example "theme" may be a popular children's'
television show, such as, for example, "In The Night Garden".
Various different "themes" can be stored as theme records 7322, for
example different sets of content for use with boys and girls, or
for children of different ages. Whatever content is used, the child
view module 7330 accesses the content, and controls the patient
display 76 to display the content at the fixation point, and at the
target points, as directed by the clinician view module 7326. In
this respect, the clinician view module 7326 may pass test plan
data to the child view module, specifying the positions at which
the target graphic should be displayed, as well as other settings
relating to the display, such as, for example, whether the fixation
graphic should be displayed at the same time as the target graphic,
or not.
[0085] Additionally provided is a patient class module 7332 which
provides methods that store all of the patient's details along with
lists of their previous tests. It also provides access to this data
as well as providing functions which work on that data, such as age
calculations. In particular the patient class module 7332 accesses
patient record data 7324, with a separate patient record being
stored for each patient that is tested, indexed by a patient ID.
The clinician view module 7326 therefore delegates the task of
managing patient records to the patient class module 7332. That is,
when the clinician view module is controlled by the clinician to
load patient details, the clinician view module passes the patient
ID received from the clinician to the patient class module 7332,
which then accesses the appropriate patient record, processes that
record, and passes back display data to the clinician view module
7326, to be displayed on the clinician display 74.
[0086] The point class module 7334 is a further software module
that contains methods that control information about each test
point in a test, including position of a test point in terms of
angle from the horizontal and vertical, and angle of magnitude.
Each test point is stored as a test point object, controlled by the
point class modules 7334. The clinician view module 7326 therefore
works with the point class module 7334 to determine at which point
the target graphic should be displayed next, and to control the
clinician view to indicate to the clinician where a point is to be
displayed. Likewise, the clinician view module passes back data
received from the clinician (i.e. yes, no, or maybe), as to whether
the point was recognised by the user, which is stored against the
point by the point class module.
[0087] Additional modules include the settings module 7336, and a
session settings module 7338. The session settings module 7338
provides a GUI to allow various settings of the application to be
set by the clinician user, which settings are then stored by the
settings module 7336. The session settings module 7338 also
contains default settings, which are used by the application under
the control of the clinician view module 7326 when it is first
loaded. During a particular session these settings may be changed,
and stored in the settings module 7336, for example to change the
application settings on a test by test basis. Various settings may
be set by the clinician user, such as, for example, the size of the
fixation graphic, the size of the target graphic, whether the "auto
add points" function is enabled (auto add points will be described
later), the interval at which points should be added by the auto
add function, whether the fixation graphic should be shown at the
same time as the target graphic, whether in the case of a video
target the video should be looped, whether random video content
should be used, whether the target graphic should be automatically
hidden after a preset amount of time of being displayed, etc.
[0088] In addition, various hardware settings may also be set,
including display resolution settings of both the patient and
clinician displays, as well as settings such as screen size of the
patient display, and the distance from the screen that the child
subject is placed. The settings module 7336 and session settings
module 7338 allow all of these settings to be altered by the
clinician.
[0089] A further module is the test module 7340. The test module
contains all of the information required for a peripheral vision
field test. This includes a list of the points tested and other
information required for calculating the results. This module is in
charge of managing the test as it is run, for example providing a
test point that still needs testing to the main interface. The auto
add function is also part of the test module 7340. The test module
7340 works closely with the point class module 7334 so as to
actually run peripheral vision tests, comprising multiple points
controlled by the point class module 7334. The test module 7340 is
therefore called by the clinician view module 7326 when a test is
to be run, and which then delegates the management of the test to
the test module 7340. The module 7340 then calls various data and
information relating to specific test points from the point class
module 7334 and passes data back to the point class module 7334
during a test. Together, therefore, the clinician view module,
point class module, and test module contain all of the data and
methods necessary to run a test.
[0090] Finally, a "template maker" module 7328 is provided. This
allows a clinician user to specify test point locations to create a
new test profile. That is, the template maker module 7328 causes a
target graticule to be displayed on the clinician display 74, and
allows the clinician to select points on the graticule which will
then form test points at which the target graphic will be
displayed. The collection of points or metadata relating thereto
can then be stored as a new test template, for use by the test
module 7340. Likewise information regarding the specific points
that make up the test are stored as point objects that are used by
the point class module 7334.
[0091] FIGS. 9, 10 and 11 show a peripheral vision testing system
of the second example set up and ready to be used. In particular,
FIG. 9 is a photograph of the testing room, with the patient
display shown on the right, and the corresponding clinician display
shown on the left. FIG. 10 illustrates a test subject placed in
front of the patient display, and FIG. 11 is a photograph of the
clinician display taken at the same moment in time as when the
photograph in FIG. 10 was taken.
[0092] Having described the hardware and software of the perimetry
measurement system of the second example, the operation of the
system in performing a test will now be described with respect to
FIGS. 12 to 19.
[0093] To operate the system the computer 72 is first booted up,
and the measurement application started. When the application
starts, it displays an input box to the clinician user, to ask for
the patient ID of the present patient to be tested to be entered.
The clinician then enters this information, and the clinician view
module 7326 asks the patient class module 7332 to access the
appropriate patient record (or to create a patient record if none
is available). The clinician view module 7326 then displays the
clinician graphical user interface 130 as shown in FIG. 13, having
the retrieved patient details 132 included in a box on the left
hand side. With respect to FIG. 12, these steps are performed at
steps 12.2 and 12.4. Therefore, after step 12.4, the clinician
display 74 displays the clinician graphical user interface 130 in
the form as shown in FIG. 13.
[0094] Next, the clinician determines at step 12.6 that a new test
is to be performed and therefore clicks the "perform the test"
button provided by the GUI 130 on the screen. This then brings up a
list of test templates, as shown by templates 134 in FIG. 14. This
is performed at step 12.8. The test templates are either
predetermined test templates provided with the application, or may
be templates which the clinician has previously created, using the
template maker module, described in more detail later. The
clinician user then selects a template to be used at step 12.10,
specifying at the same time which eyes are to be tested. Before
starting the test the clinician user should also ensure that the
child subject has been placed in front of the screen, with the
appropriate eye covered, being the opposite eye to the eye that is
to be tested.
[0095] After selecting a test template, the clinician user also
selects a theme appropriate to the child subject. As discussed
previously, several different content themes are stored, which are
appropriate for children of different sexes, and different ages.
The clinician user should select a theme appropriate to the age and
sex of the child subject. Once this has been achieved, then the
test can commence.
[0096] As in the first example described previously, within the
second example the test comprises showing the fixation target
graphic or video to the child subject at the centre of the screen,
and then intermittently showing a target graphic or video in a
different position on the screen, and determining whether the child
subject sees the target graphic. FIG. 15 illustrates the patient
display showing the fixation graphic or video 150 in the centre of
the screen. This is the "default" display, which brings the child
subject's attention back to the centre of the display. When a
target graphic or video is being shown, the display changes to that
shown, for example, in FIG. 16, where the target graphic 160 is
shown in this case above the fixation point. It will be understood
that the target graphic 160 can be shown at any position on the
screen, as defined by the test points in the test template.
[0097] FIGS. 17 and 18 illustrate the clinician view during the
test. FIG. 17 illustrates the clinician GUI 130, which has in the
main window thereof the video image captured by the camera 78, of
the child subject. In FIG. 17, the situation is that the fixation
video or graphic is being shown at the centre of the screen, and
hence it appears that the child is looking directly at the camera.
However, the clinician display is also augmented by an indicator
172, which indicates to the clinician the position where the target
graphic or video is about to be displayed to the child. The colour
of the indicator 172 indicates whether the target graphic or video
has been displayed at that position. For example, a red indicator
172 indicates to the clinician that the target graphic or video has
not yet been shown to the child at that position, but is about to
be shown to the child at that position. When the target graphic or
video is then shown to the child at the indicated position, the
colour of the indicator 172 changes, for example to green. This
situation is shown in FIG. 18. Here, in comparison to FIG. 17 it
can be seen that the colour of the indicator 172 has changed
slightly, and also that the direction of gaze of the child subject
has shifted to the direction of the position of the target graphic
or video. In such a case the clinician can be relatively certain
that the child has seen the target graphic or video, and hence the
clinician is able to select the "yes" button 174, provided by GUI
130. It should be noted that the "yes", "maybe", and "no" buttons
in the GUI only become available for selection during the period in
which the target graphic or video is being displayed to the child,
and not before. Thus, for example, in FIG. 17 the buttons are
"greyed" out, whereas in FIG. 18 they are available for
selection.
[0098] Returning to FIG. 12, the above described operations are
repeated for every point in the test. That is, to perform a test
firstly a processing loop is started at step 12.14, to ensure that
each point in a test template is displayed. For each point in the
test template at step 12.16 the clinician monitor is updated with
an indicator 172 indicating where on the patient display the target
graphic or video will be displayed. Thus, step 12.16 corresponds to
the state shown in FIG. 17, described previously. Next, at step
12.18, the target graphic or video in accordance with the chosen
theme animation is displayed at the present target point position
on the child monitor. This therefore corresponds to the states
shown in FIG. 16 for the patient display, and FIG. 18 for the
clinician display. As the target graphic or video is being
displayed, the clinician can feedback using the buttons 174 as to
whether the child has seen the point. How the system next operates
then depends on the clinician feedback.
[0099] More particularly, if the clinician selects the "yes" button
174, as shown in step 12.22, then the positive result is logged by
the test module 7340, against the particular test point. Then, the
next point can be processed, and the processing loop returns to
step 12.16, for the next point. Alternatively, if the feedback is
"maybe", obtained by the clinician selecting the "maybe" button,
then that point is maintained in the test list, and is, for
example, put as the last point in the list, so that it is tested
again before the test ends. Alternatively, the point may be placed
at a random point in the remaining test point list, so that it does
not necessarily appear last, but does appear subsequently. The
point can then be tested again later in the test until a definitive
answer is obtained from the clinician. In further examples it may
be possible to keep a track of the number of times that a point has
"maybe" recorded against it, such that it is only placed back into
the test list a certain number of times such as, for example, once,
twice, or three times at the maximum.
[0100] Regarding the ordering of test points in a test, preferably
the display order of test points is such that the test subject is
not able to discern a particular pattern or order to the test
points, such that subject has some indication of where the next
test point will be displayed based on those shown previously.
Instead, a random or apparently random order of test points may be
used.
[0101] If the clinician feedback is that the target graphic or
video was not seen at the test point position, then again this
result is logged at step 12.28. According to the settings of the
test the test may then automatically add a point into the test,
using an "auto add feature", to be described. If a test point is
added in using the "auto add feature" then the new test point is
added into the list, and although it may not be the next point to
be tested, it will be tested before the test completes. When a
point has been auto added into the test, processing then proceeds
back to step 12.32, wherein it is determined whether there is a
next point to test. Where a point has been auto added in to the
test at step 12.30, there will always be a next point to test.
Likewise, where a point has been kept in the list at step 12.38,
similarly the test will continue.
[0102] Eventually, all of the test points will have been tested,
and "yes" or "no" results stored against each point. Thereafter,
the patient test results can be stored in the patient record, at
step 12.34, and subsequently viewed by the clinician, as will be
described later.
[0103] Turning first to FIG. 19, however, the operation of the
"auto add" feature of step 12.30 will next be described.
[0104] FIG. 19 shows an example result plot of points that have
been tested, to give a peripheral vision field plot, in this case
for a subject's right eye. Result plot 190 comprises the positions
of points that were tested 194, overlaid on a graticule 196. Each
test point 194 is characterised by two values, firstly an angular
value from either the vertical or horizontal lines of the plot, and
secondly a magnitude value, also in degrees, indicating angle of
gaze from the centre point out to the point. Thus, for example, a
test point may have a position of 45.degree., being the rotational
angle from the horizontal, or the vertical, and a magnitude of
50.degree., being the angle of view from the centre point, for a
subject positioned in the centre.
[0105] The auto add function makes use of this definition of test
points so as to add points into a test, the added points being of
reduced magnitude, when a test point has not been spotted by a
subject. Thus, for example, as shown in FIG. 19, an original test
point was test point 1982, which has an angular rotational position
of approximately 150.degree. (assuming 0.degree. is the positive X
axis) and an angular magnitude of 50.degree. (it is positioned on
the 50.degree. graticule line). In the event that this test point
is not spotted by the test subject, and the auto add feature is
enabled, an extra test point 1984 is added into the test template,
having an identical rotational angular position, but a different
angular magnitude. For example, in this case the angular magnitude
has been reduced by 5.degree., so that the magnitude of point 1984
is 45.degree.. This is then included in the test template, and a
test graphic is in due course presented to the test subject at this
position.
[0106] In this case, however, the test subject has still not
spotted the test graphic at this position, and hence the auto add
feature has added in a third test point 1986 of even small angular
magnitude, but having the same rotational angular value. Thus, test
point 1986 is of angular magnitude 40.degree.. In this respect, it
can be seen that the auto add feature has reduced the angular
magnitude of each point by 5.degree. each time. However, this step
size can be set by the clinician user. In the case of FIG. 19, the
second added test point 1986 has still not been spotted by the user
when the test graphic is displayed thereat, and hence a third test
point is added, being test point 1990. This has an even smaller
angular magnitude of 35.degree.. In this case, however, when the
test graphic is presented to the user at this position, the user
has seen the test graphic, such that it is not then necessary to
add in further test points.
[0107] However, in this example, the auto add feature has brought
to light that the test subject has significant visual field loss in
the lower left quadrant of his right eye, which should be
investigated further.
[0108] Regarding the display order of points that have been added
in to a test, as noted earlier an added point may not necessarily
be the next test point to be displayed. Whilst in one embodiment an
added test point at the same rotational angular position but
smaller angular magnitude as a missed point may be displayed in
order after the missed point i.e. it is the next point to be
displayed after the missed point, in other embodiments an added
point may be displayed later in the test series, after other points
at different rotational angular positions have been displayed. By
so doing, the test subject's attention is drawn away from the
particular rotational angle region in which the missed test
point(s) and any consequentially added test point(s) are located,
and the test subject does not become alerted or suspicious to any
unusually long gap in test images being displayed, such that their
eyes may wander from the fixation image.
[0109] More particularly, imagine a test subject is viewing the
fixation image at the fixation image position such as the center of
the screen. If a point is then displayed that the test subject does
not see, then a point will be added in at the same rotational
angular position, but of lesser angular magnitude. However,
depending on the time taken the test subject may by now have
expected to see a test image, and may start to look around the test
area, which would spoil the test. Therefore, in order to stop this,
a test image at a different rotational angular position can be
displayed, for example at a test position that the test subject has
already been positively tested at i.e. at a location where the test
subject saw the test image. Alternatively, another test point can
be displayed to the subject, from the set of test points to be
tested that have not yet been shown to the user. As such, the test
subject is not alerted to the fact that a test point had to be
added in, and the continuity of the test is maintained from the
test subject view point.
[0110] In such a case, the point that has been added in is then
displayed later in the test sequence, after one or more other test
images at different rotational angles have been displayed to the
test subject. If multiple points have to be added in at the same
rotational angle, due to the reason that the test subject keeps
missing the images (as in FIG. 19 and described above), then the
same procedure as above may be used i.e. another test image at a
different rotational angular position is displayed before the
added-in test point is displayed at the problematic angular
position where the user did not see the test image.
[0111] Turning now to how the test results may be reviewed, the
clinician view module provides the ability to review a test that
has just been performed, together with previous tests. In addition,
the tests may be printed. FIG. 20 illustrates the steps involved in
displaying test results. Firstly, at step 20.2 a patient ID is
entered, and at step 20.4 the patient records are retrieved, and
displayed. In this case, as shown in FIG. 22, if the patient has
had a test performed then the results of that test will be stored
with the patient record, and retrieved with the patient record for
display to the clinician. FIG. 22 illustrates how the results of a
test 222 are displayed to the clinician, as test plots 224, in the
main clinician view. The test plots 224 can be displayed all at
once, as shown, or may be displayed individually for each eye, or
for binocular vision, by appropriate selection of check boxes 226.
In addition, the plots may use bezier curves, as shown, to connect
the points, or alternatively may, although perhaps less helpfully,
connect the test points with straight lines, by the clinician
ticking the "straight line" check box 228 in the GUI. In order to
print the results, at step 20.10, the clinician may select the
print icon in the top left of the GUI, and then the selected
graphical records are printed.
[0112] It is also possible to show multiple results, if multiple
test results are available. The multiple results can be overlaid on
top of each other for comparison, as shown in FIG. 23. In FIG. 23
it can be seen that two previous sets of test results are available
in column 222, and that these are shown as overlaid test results
230. This view is useful for showing changes in peripheral vision
fields. For example, in FIG. 23 it can be seen that the peripheral
vision field of the subject's left eye has deteriorated markedly in
the upper right quadrant. Multiple tests can be overlaid one on top
of each other, and different colour plots are typically used to
identify which test is which.
[0113] FIG. 24 illustrates how only a subset of multiple tests may
be displayed. In particular, FIG. 24 illustrates multiple left eye
plots, allowing the clinician to see in more detail the change in
peripheral vision field in the upper right quadrant of the
subject's left eye. In addition, here it can be seen that the auto
add feature has been used to add in points on the second test as
shown by the sequence of points 242 shown on the results.
[0114] In addition to the above described features, preferably the
viewing feature also allows the clinician to zoom in to parts of
the plot, and to scan and pan around the graphical result.
[0115] As discussed previously, the measurement system also
includes a template maker module, to allow a clinician to make new
test templates, being sets of test points. This allows a clinician
to design a test to try and focus on particular areas of a
subject's peripheral vision. In order to produce a new test
template then, as shown in FIG. 21, at step 21.2 the clinician
selects a new template to be produced, using the "template maker"
button in the GUI 139. This causes the new template maker graticule
to be displayed at step 21.4, as shown in FIG. 25. The template
maker graticule displays in a new window 250. Using a mouse pointer
the clinician then marks on the template maker graticule the
position of test points to be tested. As shown in FIG. 25, two
points 252 have already been marked by the clinician, and the mouse
cursor is shown as point 254. The template maker graticule can have
a "snap" feature, to cause the cursor to snap to multiples of
1.degree., 5.degree., 10.degree., 15.degree., 20.degree., etc for
both rotational angle and magnitude of the point. Alternatively,
the clinician may place points wherever he or she wishes within the
template maker graticule. To set a point the clinician simply moves
the cursor 254 to the desired point, and clicks the mouse. The
position is then stored as a new test point. Once the clinician has
finished, and sufficient test points have been set, the set of
points may be stored as a new test template, at step 21.8. That
test template can then be selected for use at, for example, step
12.10 of the process of FIG. 12, described previously.
[0116] The provision of the template maker therefore allows a
clinician to define tests to try and make a specific diagnosis or
test particular areas of a subject's peripheral vision field in a
simple and user friendly manner.
[0117] Overall, therefore, the peripheral vision measurement system
of the second example provides a reliable and simple to use
interface for performing peripheral vision tests, and allows for
the easy management and display of test results. Testing of the
system on live subjects has shown the reliability and feasibility
of the system.
[0118] Various modifications may be made to the above described
examples to provide further examples. For example, within the two
examples described above a television type display screen such as a
plasma, LCD, or LED display is used as the patient display.
Alternatively, in other embodiments, a projection type display may
be used. In this case, a video projector is used to display the
fixation and target images onto a screen. A small hole may be cut
in the screen at the centre, allowing the camera to be mounted
behind the screen. This has advantages in that the camera is then
not so prominent to the subject, and does not occlude any of the
patient display screen. In addition, using a projection type system
allows a larger image to be obtained than is the case with a
display of fixed size such as a plasma, LCD, or LED screen.
However, of course, where the size of the display changes due to
projection characteristics, the positioning of the subject in terms
of how far away from the display they are sat must be carefully
made. The distance of the subject from the screen is a parameter
that can be entered into the settings module of the application, to
allow for automatic adjustment of target graphic and video
positions.
[0119] In the above examples, it is described that the target
graphic is preferably a video, of a theme selected by the clinician
so as to be attractive to the child subject. One particularly
preferable feature of using a video is that the same video can be
used for both the fixation graphic, and the target graphic, without
interruption. That is, the child subject starts watching a
television program as part of the video content displayed at the
fixation point, and then that video content is then continued in
the target video at the target point position in a seamless manner.
Thereafter the video is then continued at the fixation target when
it is next displayed, followed by then being displayed at the next
target point, and so on. Thus, as far as the child subject is
concerned, all that happens through the test is that the position
of the video on the screen changes, but that the story contained
within the content is continued.
[0120] There are also several variations as to whether the fixation
target is displayed at the same time as the target graphic or
video. For example, it may be that the fixation target is a video,
and is displayed constantly, with the target graphic then being
either another video, or some other animated graphic to try and
catch the child's attention away from the fixation video. However,
it is thought that better results are obtained by removing the
fixation video from the display screen when the target graphic is
being shown. For some subjects, however, and particularly those who
have performed the test before, removal of the fixation video may
be an indicator that they should scan the rest of the screen to try
and locate a target video. Therefore, to try and overcome this, for
some tests it may be preferable to maintain the fixation target on
the screen at the same time as the target graphic or video is being
shown, so as to try and maintain the subject's attention on the
fixation point, and thereby achieve better results.
[0121] In one example the fixation image may freeze at the last
video frame displayed before the target image is displayed, and the
freeze frame then continues to be displayed whilst the target image
is displayed. The target image in such an example may be the video
image (that then continues to play), or another image. Where
another image is used the video soundtrack may continue to play,
even though the video image to match the soundtrack is not being
shown as either the fixation image, or the target image.
[0122] It will therefore be appreciated that there are several
combinations of options as to format of the fixation images and the
target images. The settings module provides settings to allow the
clinician to chose any of the above options, depending on the
subject.
[0123] Within the above first and second example the clinician
display is augmented with indicators as to where the target graphic
is presently been shown on the patient display, and also as to
where the target graphic may next be shown. In the above example a
colour coding system is used to distinguish whether the indicator
shows that the target graphic is presently being displayed at the
indicator location, or whether it is about to be displayed. In
other examples instead of the colour of the indicator changing or
otherwise being indicative, some other characteristic of the
indicator is altered, for example its shape. For example, a square
indicator may be used to show the position of where the target
graphic is about to be displayed, and the shape of the indicator
then changes to circular when the graphic is being displayed. In
further examples two or more characteristics of the indicators may
be altered, to try and convey more information. For example, both
shape and colour may be used, with an indicator of a first shape
illustrating a test point that has already been shown to the
subject, and an indicator of a second shape illustrating test
points that are about to be shown, or that are still to be shown in
the future. The colour of the indicators of the first shape may
indicate whether the point was seen by the user or nor. For
example, a first colour may indicate that the point was seen, and a
second colour may indicate that it was not. A third colour may be
used to indicate that the point was a "maybe".
[0124] Within the above examples the images of the test subject are
displayed to the clinician user throughout the test, and the
clinician user makes a decision and provides feedback as to whether
the subject saw a particular target point when it was displayed. In
another example, however, instead of (or in addition to) the
clinician reviewing the video images of the subject during the test
the video images of the subject may be recorded for review by the
clinician later. That is, whether displayed to the clinician in
real time or not, the video images of the test subject performing
the test may be stored, so that they may be reviewed by a clinician
at a later time. In this respect, the video data is typically
stored with the patient record.
[0125] Where the video images of the test subject are not to be
shown to the clinician in real-time, then in another example the
same display device (HDTV, computer monitor, etc.) may be used to
display the video images to the clinician as was used to display
the test images to the subject. Hence, some equipment re-use can be
obtained.
[0126] In another example, where the video images of the test are
recorded a clinician can then review the test video later and make
decisions at that moment in time as to whether a child saw a target
graphic or not. In this case as the video of the test is played
back to the clinician user the clinician video display is
preferably augmented with the indicators to show where the test
graphic is about to be and then subsequently is being shown on the
patient display, so that the clinician has the same experience as
if he or she was conducting the test live. Therefore, as well as
storing the video image data test meta data synchronised with the
video data is also stored that instructs the clinician view module
to augment the clinician video display with target point position
indicators even when the video being shown is recorded video.
[0127] The above example allows a clinician to review a test at a
later time, and hence allows tests to be performed even when a
clinician is not available at the time the test is to be performed.
However, because the clinician is not available to make a decision
as to whether a target point was seen during the test, then the
"auto-add" feature previously described becomes less useful.
Instead, tests can be designed with test points which display
target graphics or videos across the whole field of the view of the
test subject, rather than trying to restrict test points to the
edge of the vision field, and then adaptively adding test points
in.
[0128] In other examples it will also be understood that, whether
based on recorded test imagery or live test imagery, the clinician
need not be co-located with the test subject, and that the
clinician may be located at a separate display located in another
room, building, city, country, or even continent, with the video
imagery of the test subject then being provided to the clinician's
computer over a network. In this case typically two control
computers are required, being one to operate the patient display,
and one to operate the clinician display. One of the computers may
operate in a server mode, and perform most of the processing,
controlling the client terminal to produce the appropriate display.
For example, the computer at the clinician end may contain most of
the software, and store test results, patient records, and the
like, and control the patient display remotely by virtue of a
relatively dumb client installed on a computer co-located with the
patient display.
[0129] In another example, the application may be completely
network based, and run on a network server which is not located
with either the patient display or clinician display. In this case,
the patient display and clinician displays can run relatively
lightweight dumb clients, with most of the processing then
performed over a network by the network server. Patient records,
test results, test video, and the like can be stored at the network
server, or alternatively elsewhere on another storage server to
which the network server running the testing system has access.
[0130] In the above first and second example the clinician is
relied upon to review the video imagery of the test subject and
decide if the subject's direction of gaze moves towards a test
point when the target graphic or video is displayed on the patient
screen at that location. However, in a further example this
decision making process of the clinician may be replaced by an
image processing function, such as, for example, the template
matching techniques described by Suga et al, supra. In such an
example the video images of the test subject are processed by the
template matching image processing algorithm to determine if the
subject's direction of gaze moves towards a target graphic or video
when displayed, and the results then recorded automatically. In
this example, there is no need to provide the clinician display,
although such a display may be provided for monitoring purposes. In
addition, if the image processing algorithm determines that a
target is not seen then the auto-add function described above can
be used to add in an easier test point, which can then be presented
to the test subject. In this respect, the image processing
algorithm replaces the receipt of the clinician feedback of step
12.20 of FIG. 12, but in all other respect the processing and
operation can remain identical. For example, continuous video may
be used for the target and fixation points.
[0131] The auto add function is particularly advantageous, and can
be used where the decision as to whether a target was seen by the
subject at a test point is made either by the clinician user, or
using image processing techniques. In particular the auto add
function allows the location of the test points to be adapted
dynamically during the test, such that the most accurate plot of a
subject's visual field can be obtained as possible. In particular,
by incrementally adjusting the angular magnitudes of the test
points, so as to reduce the angular magnitude, the extent of a
subject's visual field can be adaptively explored during a test, so
as to determine the visual field extent.
[0132] In another example, where an automatic gaze tracking
technique is used, such as in Suga et al, in order to supplement
the image processing algorithm that is used to track the eye
position the head position can also be independently tracked using
natural or other face-attached features to resolve head position.
For example, a fiducial, calibration sticker may be placed on the
eye patch (which is already worn by the subject when one eye at a
time is being tested). One problem in automated eye tracking
techniques is to resolve the subject head position in the first
place, from which the eye position and direction of gaze can then
be determined. In other words,
eye-in-space=eye-in-head+head-in-space. By using a known
calibration symbol such as a cross or some other easy-to-recognise
symbol displayed to the camera on the outer surface of the eye
patch, head position can be readily resolved in an image by looking
for the known symbol shape. Once head position has been established
in the image, then identification of eye position within the image
becomes more straightforward. That is, spatial segmentation of the
images to be processed becomes easier, as there is a generally
known a priori relationship between the position of the calibration
symbol, and the head orientation and position of the test
subject.
[0133] In a further example where image processing to provide
automatic gaze tracking in used, as well as providing spatial
segmentation of an image using a known calibration symbol as
described above, in addition or alternatively it is also possible
to temporally segment a video stream, to reduce the number of video
images that need to be processed. More particularly, because the
system knows precisely when target images will be displayed to the
test subject (since the system generates and controls display of
the target images), then it only becomes necessary to image process
the few video frames before and after the target image event
(probably at most 1.0 sec=0.5 sec+0.5 sec of footage each side) to
determine eye gaze direction of the subject. At other times the
direction of gaze of the subject is generally irrelevant.
[0134] One caveat to the above is that the subject should
preferably be viewing the fixation image at the time the target
image is to be displayed. Only if this is the case is an accurate
test of peripheral vision performed, as if the subject is looking
elsewhere on the screen when a target image is shown the relative
position of the target image in the subject's field of view will be
different than the case where the subject if looking at the
fixation image. Therefore, in another example eye gaze
determination is performed in advance of a target image being
shown, to determine whether the subject is looking at the fixation
image, or at another part of the screen. Display of the target
image is then adapted in dependence on the determination as to
whether the subject is looking at the fixation image, in that the
display may be delayed until the subject's gaze returns to the
fixation image. In this way target images are shown when the
subject is looking at the fixation image, and hence test accuracy
can be improved.
[0135] Within the above example the determination as to whether the
test subject is looking at the fixation image may be performed
automatically, for example by an image processing algorithm such as
Suga et al, or by the clinician. In the latter case, the clinician
GUI may further comprise a button which is activated by the
clinician during the test to indicate to the system that the test
subject is looking at the fixation image before a target image is
to be displayed. Once the clinician activates the button the system
then proceeds to display the target image at the indicated target
location.
[0136] In another example of the invention an automated eye
tracking algorithm such as Suga et al can be combined in hybrid use
with the augmented clinician view in a semi-automated fashion. For
instance, an automated analysis of the 1 sec eye movement clips per
target mentioned above can be done to enhance accuracy by catching
false positives and negatives, either as: a) automatically for each
target after it appears (e.g. giving percentage confidence) during
the course of a manual, augmented-view test run by the clinician;
b) in batch for the full test sequence at the end of the
augmented-view test run by the clinician; or c) automatically (as
in Suga), instead of an augmented-view clinician test but followed
with full or partial manual augmented-view human-based testing by
the clinician. The latter human confirmation will be very rapid as
it will merely run through the 1 sec (or shorter) clips, or even
organise them usefully in a combined full-screen view pointing to
the targets of most interest to review for accuracy. Such a human
review could also be done during the course of the test, if the
system is allowed to function automatically for some of the time
during the process.
[0137] Thus, even where an automated gaze-tracking system is used
to perform peripheral vision tests, such as in Suga et al, such an
automated system can be complemented by the augmented clinician
view of the previous examples described above, to allow the
clinician to check the results of the automated system, and to
increase confidence levels in each result. Alternatively, where, as
in the first and second main examples described above, the
clinician is the primary determiner of whether a target image was
seen, an automated image processing system may be used as back up
to check the clinician determined results, and increase the
confidence in each test result. As noted, this image processing may
be performed as a test progresses, or in batch at the end of the
test, preferably by processing the time segmented video stream, so
as to process only those video frames around the time the target
images were displayed.
[0138] Various further modifications, whether by way of addition,
deletion, or substitution will be apparent, to the intended reader,
being a person skilled in the art, to provide further examples, any
and all of which are intended to fall within the appended
claims.
* * * * *