 |
Description  |
|
|
CROSS REFERENCE TO RELATED APPLICATION
This application relates to copending application Ser. No. 557,263, filed
Jul. 24, 1990, assigned to the same assignee as the present invention.
BACKGROUND OF THE INVENTION
This invention relates to a system and method for aligning the eyes of a
subject.
There are a number of situations where the alignment of a subject's eyes
are important. The subject may be human or may be a research animal.
When taking certain types of eye measurements, one may need to know that
the eye alignment is in a particular reference position. For example, if
one measures the cornea of a person's eye before some type of treatment
and one wishes to repeat those measurements after the treatment to
determine how much, if any, the treatment has affected the measurements,
one must ensure that the eye alignment is in the same position each time
the particular measurements are made. Otherwise, the difference in data
from before and after the treatment might be due to a change in eye
alignment rather than the treatment.
In addition to those situations where one needs to ensure that eye
alignment is in the same position for two or more different measurements,
there are situations where eye alignment is desirable for diagnostic
measurements of eye performance. There are situations when a human subject
can simply be requested to fixate on a particular object. Thus, the human
may state that he is currently looking at a light source, thereby
providing "subjective" eye alignment information. However, there are
situations where a doctor or researcher would like "objective" eye
alignment information indicating the orientation of the eye and, to the
extent possible, indicating what the eye is viewing. For example, very
young children cannot be relied upon to fixate on such an object for
measurements, such as refraction measurements which are very desirable to
ensure that "in focus" images are being received when the child's brain is
learning to interpret images. Likewise, adults subjected to extended eye
examinations may become tired or subject to other duress and fail to
maintain reliable fixation. A patient who is subjected to a therapeutic
process such as laser ablation eye surgery may not be able to maintain
desired eye orientation over an extended treatment time because of applied
anesthesia, fatigue, or distraction by the procedure. Further, a research
animal usually cannot be trained to fixate during eye measurements.
In each of the above cases, the failure or inability of the subject to
maintain eye fixation upon an object can produce eye measurements or
treatments that are seriously in error.
Therefore, there are situations where absolute eye alignment data is needed
(i.e., the eye is aligned in a certain manner) and situations where
comparative eye alignment data (i.e., the eye is in the same alignment as
when earlier measurements were taken) are needed and one cannot rely upon
a subject maintaining the alignment.
The use of various optical/electronic systems for making eye measurements
has become more common in recent years. One example of such a system is
shown in U.S. Pat. No. 4,213,678 to Pomerantzeff and Webb issued on Jul.
22, 1980. That patent discloses a scanning laser ophthalmoscope (SLO) for
scanning a portion of the eye fundus. That patent, coinvented by one of
the present inventors, is hereby incorporated by reference.
OBJECTS AND SUMMARY OF THE INVENTION
Accordingly, a primary object of the present invention is to provide a new
and improved method and system for determining objective eye alignment.
A more specific object of the present invention is to provide objective eye
alignment information which may provide the eye alignment relative to some
external point or line, such as an instrument axis, or to provide eye
alignment information by indicating when the present eye alignment is the
same as an earlier eye alignment.
A further object of the present invention is to provide eye alignment data
which can be used in combination with other diagnostic and therapeutic
instruments. In other words, an object is to allow eye alignment data to
be provided without requiring instrumentation which would block or prevent
use of diagnostic and/or therapeutic devices which may advantageously use
eye alignment information for improving their diagnostic and/or
therapeutic operations.
Yet another object of the present invention is to allow one to bring the
eye of a subject to a particular desired alignment, while the subject is
under general anesthesia or is otherwise unable to cooperate in bringing
his eye to a particular alignment.
A still further object of the present invention is to provide eye alignment
information in a reliable and convenient fashion.
A more specific object of the present invention is to provide an imaging
system and a method of use of an imaging system for determining objective
eye alignment.
The above and other objects of the present invention, which will become
more apparent as the following detailed description is considered, are
realized by a method of determining objective eye alignment. A subject's
eye is positioned relative to an imaging system to allow imaging of the
eye. The imaging system provides an iris image of an iris portion of the
subject's eye and a retinal image of a retinal portion of the subject's
eye. The iris image and the retinal image are used to determine objective
eye alignment. Preferably, the imaging system obtains the iris image and
the retinal image within a time no greater than 50 milliseconds. To
minimize intervening eye movements the time of obtaining the images would
preferably be substantially less than that. The iris image and the retinal
image are simultaneously displayed on a first monitor of the imaging
system. (As used herein, "simultaneous display" means that the images
appear simultaneous to a human eye.)
The eye alignment may be determined in two different manners. A first way
of determining the eye alignment includes the step of comparing the iris
image and the retinal image with a stored image of the iris portion and
the retinal portion, the stored image corresponding to eye alignment at a
previous time. More specifically, this first manner involves finding an
eye position at which the iris image and the retinal image are
substantially identical to the stored image. A second manner of
determining eye alignment involves finding a relationship between an
instrument axis of the imaging system and an axis of the subject's eye.
More specifically, the axis of the subject's eye may be the pupillary
axis.
The iris image and the retinal image are provided by the imaging system
scanning a beam across the eye and detecting scatter from the application
of the beam to the eye. The beam is preferably a beam of electromagnetic
radiation in the infra-red, visible or ultraviolet portions of the
spectrum..
The method of the present invention may alternately be described as a
method of determining objective eye alignment including a step of scanning
a light beam from an imaging system across a subject's eyes such that the
light beam strikes an iris portion adjacent a pupil and strikes a retinal
portion. Scattered light from the application of the light beam to the
iris portion and the retinal portion is then detected by operation of the
imaging system. Nearly simultaneous (meaning simultaneous or within 50
milliseconds thereof) views of the iris portion and the retinal portion
are generated by operation of the imaging system. The nearly simultaneous
views are then used to determine objective eye alignment. The light beam
is scanned between extremes of scan which cross in front of the object's
eye. More specifically, the extremes of scan preferably cross at least
three centimeters in front of the subject's eye (i.e., the closest point
on the subject's eye to the scan crossing point is 3 centimeters or
greater). The light beam applied to the subject's eye has a diameter of no
more than 1 millimeter at the iris.
The system of the present invention is a system for measuring objective eye
alignment including a scanner for scanning a beam across a subject's eye
such that the beam strikes an iris portion adjacent a pupil and strikes a
retinal portion. A detector for detecting scatter from the application of
the beam to the iris portion and the retinal portion is operatively
connected to a first monitor for showing simultaneous views (i.e., they
appear simultaneous to a human eye) of the iris portion and the retinal
portion based upon the scatter detected by the detector. The simultaneous
views provide objective eye alignment information. More specifically, the
scanner scans a light beam. The system may further include a second
monitor and storage means to supply the second monitor with a stored image
of the iris portion and the retinal portion at a previous time.
BRIEF DESCRIPTION OF THE DRAWINGS
The above and other features of the present invention will be more readily
understood when the following detailed description is considered in
conjunction with the accompanying drawings wherein like characters
represent like parts throughout the several views and in which:
FIG. 1 shows a general schematic diagram illustrating how the present eye
alignment system may be used in conjunction with another eye diagnostic
instrument;
FIG. 2 shows a simplified view illustrating how the present alignment
system may apply beams to the eye of a subject;
FIG. 3A, 3B, and 3C show separate images which the present system may
provide for purposes of determining eye alignment information;
FIG. 4 shows an arrangement wherein the present eye alignment system is
operating in conjunction with a known instrument which provides
measurements of a cornea of the subject's eye; and
FIG. 5 shows a detailed view of an arrangement which may be used to provide
the scanning.
DETAILED DESCRIPTION
With reference now to FIG. 1, there is shown a simplified view of part of a
subject's eye 10 with an alignment imaging system 12 positioned to allow
determination of the eye 10. The alignment imaging system or imager 12
according to the present invention is used to provide eye alignment
information in a manner discussed in more detail below. For purposes of an
initial general explanation, the arrangement of FIG. 1 shows how this
alignment imaging system 12 may be used in combination with some other
previously known eye diagnostic instrument 14. In particular, the
instrument 14 might be some instrument which may diagnose conditions in
the eye 10. The instrument 14 may alternately be some sort of therapeutic
instrument (not separately indicated in the diagram) which may be used at
the same time as the eye alignment imaging system 12. At any rate, the
instrument 14 might be useful independently of any alignment information
provided by the alignment system 12 or, more likely, the instrument 14 may
use the alignment information from alignment system 12 for more accurately
diagnosing and/or applying treatment to the eye 10.
As shown in FIG. 1, a light source 16 may supply light to a reflector 18
which in turn supplies the light to a beam splitter 20 by way of a lens
22. The light from source 16 is applied to the eye 10 and reflects back
through the beam splitter 20 to the instrument 14. The eye diagnostic
instrument 14 may produce data about the eye 10 depending upon the light
reaching the diagnostic instrument 14.
At the same time as the instrument 14 is operating to provide the
diagnostic information, the alignment imager 12 uses a reflector 24 and
the beam splitter 20 to obtain alignment data about the eye. A manner in
which the alignment imager 12 and eye diagnostic instrument 14 may be
operated at the same time without one instrument blocking the other is
shown in FIG. 1, it being readily appreciated that various other optical
configurations could be used. A technique by which light from the
alignment imager 12 and light from source 16 used with the instrument 14
can be used at the same time without interfering with each other will be
discussed in more detail below, it simply being noted at present that
these lights may be at different wavelengths (i.e., different colors) to
avoid such interference. It should also be noted that, although the
preferred embodiment of the present invention does use an alignment
imaging system having a scanned light beam, other imaging systems might
possibly be used. Further, the instrument 14 does not necessarily have to
be an eye diagnostic instrument using a light source such as source 16.
Finally, although FIG. 1 shows the alignment imaging system 12 being used
in conjunction with the instrument 14, the alignment imaging system 12 may
be used without regard to any other instrument such as instrument 14.
The manner in which the alignment imager 12 functions to determine the
alignment of the eye will be explained by reference to the simplified
concept drawing of FIG. 2. As shown in FIG. 2, a scanner 26 scans a beam
of light 28 across the eye 10, the beam of light 28 passing through lens
30 prior to its application to the eye 10. As shown, the beam 28 has first
and second scan extremes 32F and 32S respectively, which cross at point
34, which point is in front of the subject's eye 10. By crossing in front
of the subject's eye, the extremes of scan of beam 28 will allow it to
fall upon the edges of iris 36 and, where the beam passes through the
pupil 38 within the iris 36, a retinal portion 40 at the back of eye 10.
By having the light beam 28 scan across the eye 10 and by using a detector
(not shown in FIG. 2, will be discussed in more detail below), the general
principle of FIG. 2 may be used to provide a flying spot camera. In order
to allow one to have a flying spot camera from the arrangement of FIG. 2
wherein one is able to provide an iris image corresponding to that portion
of iris 36 upon which the beam 28 hits and a retinal image corresponding
to retinal portion 40, the point 34 at which the extremes of scan cross is
typically about 3 centimeters in front of the subject's eye 10 (i.e., the
point 34 is about 3 centimeters from the closest point on the eye 10).
Additionally, and in order to provide a relatively good focus on the iris
portion, the beam 28 will have a beam diameter of no more than 1
millimeter and preferably 0.5 millimeters or slightly above that. In other
words, the beam diameter will be at or between 1.0 millimeters and 0.5
millimeters. Note that use of too small a beam diameter would degrade the
resolution at the retina 40, whereas use of too large a beam diameter
would interfere with the ability of the system to get a good focus on the
portions of iris 36 which are around the pupil 38.
In order to fully define the orientation of a generally spherical object
such as an eye, certain information is required. If one is using alignment
to simply mean the direction of a particular axis of the eye, simply
identifying the location of two points along such an axis would indicate
the alignment of the eye. However, as used herein, "alignment" will be
interpreted as also including information about rotation of the eye about
such an axis. Thus, if the particular axis within the eye corresponds to
the look direction, rotating the eye about such a look direction axis may
maintain the same object in view, even though the orientation of the eye
relative to that object has changed. (In practice, the axis within the eye
which is used for reference purposes may or may not correspond to the look
direction, but this will be discussed in more detail below.) If one wants
alignment information to include not only the direction and position of an
eye axis but also the rotational position of the eye relative to that
axis, this information can be determined by using at least three
well-separated points lying in two nearly perpendicular planes. For such a
measurement, two of these three points are used to determine the direction
and position of the axis, whereas the third point is within a plane normal
to the axis, which plane also includes one of the other two points.
The above discussion of eye alignment information will be more readily
understood by reference to FIGS. 3A, 3B, and 3C which show images
corresponding to the operation of the present invention. In particular,
FIG. 3A shows an iris image 36I having an edge 36E which defines the
outside boundaries of the pupil (not separately labeled in these figures).
For ease of illustration, the image 36I of the figures is simplified and
does not include the large number of irregularities which would be
disposed at the image of iris edge 36E. Within the closed loop of edge 36E
is the image of the retinal portion corresponding to retinal portion 40 of
FIG. 2. In particular, and as shown in FIGS. 3A, 3B, and 3C, this retinal
image part of the combined iris portion/retinal portion image includes the
fovea 42 and the optical disc 44, both of which are disposed at the retina
of the subject's eye.
Although various axes or lines might be used within a subject's eye for
defining the alignment of the eye, the discussion which follows makes
reference to the pupillary axis, which may be defined as a line between
the centroid 46 of the pupil (area within edge 36E in the image figures)
and the fovea 42 (more specifically, the center of the foveola). In FIGS.
3A, 3B and 3C, the centroid of the pupil is shown as an "X" 46.
In the view of FIG. 3A, the dextral eye shown therein with the centroid 46
of the pupil disposed at about the mid point between the fovea 42 and the
center of optical disc 44. Since the fovea and disc are separated by about
15.degree. in the vision field, the pupillary axis between the centroid
and the fovea 42 (more specifically, the center of the foveola, not
separately shown) is directed about 7.5.degree. to the right of the
alignment instrument axis. The alignment instrument axis (not separately
shown in these figures) has an axis normal to the plane of view of FIGS.
3A, 3B, and 3C. This instrument axis will be considered to extend through
the centroid 46 in order to clarify the discussion which follows. Relative
positioning between the subject's eye and the eye alignment imaging system
or other techniques may be used to provide that the alignment instrument
axis extends through the centroid, although it is not essential that a
uniquely defined alignment instrument axis extend through the centroid 46.
Instead, one might simply define an instrument point of view axis
corresponding to the direction of "view" of the instrument and which
passes through the centroid 46.
Although the pupillary axis would be directed about 7.5.degree. to the
right of the alignment instrument axis (or alignment point of view axis)
in the position of FIG. 3A, one can say that the eye is looking about
7.5.degree. to the right of the instrument axis only to the extent that
one is certain that the subject's visual axis necessarily extends between
the centroid 46 and the fovea 42 (more specifically, center of foveola).
For certain purposes, it may be reasonable or necessary to assume that the
visual axis of the subject's eye is the same as the pupillary axis.
However, the use of the present method does not require that this
assumption be made.
In the view of FIG. 3B, the eye orientation has changed so that the
pupillary axis is now only a few degrees to the right of the instrument
axis. If the fovea 42 were moved slightly to the left in the view of FIG.
3B such that it was centered beneath the centroid 46, then the pupillary
axis would be directly in line with the alignment instrument axis (or
instrument point of view axis).
In the view of FIG. 3C, the eye has rotated about an axis extending through
centroid 46, but the separation between centroid 46 and fovea 42 (i.e.,
the separation distance as viewed) has remained constant. As shown in the
views of FIGS. 3A, 3B, and 3C, the optical disc 44 may have a number of
visible fine blood vessels 48 extending near it. The vessels may include
crossing points 48C (only one shown for simplicity). By noting that the
crossing point 48C has moved counter-clockwise about 30.degree. between
the positions of FIGS. 3B and 3C, this indicates that the eye has rotated
about 30.degree. counter-clockwise. The centroid 46 may be used as the
origin for defining angles of rotation.
The alignment of the eye can thus be established by knowing the
relationship between the centroid 46 and the fovea 42. This relationship
can establish the angle between the pupillary axis and the instrument
axis. However, if the fovea 42 and centroid 46 are both in line with the
instrument axis, the position of an additional point must be used in order
to determine uniquely or establish alignment of the eye provided that we
use the term "alignment" to include rotation about the pupillary axis.
Accordingly, use of the image of the optical disc 44 and, especially, the
crossings of visible fine blood vessels such as crossing 44C will allow us
to uniquely establish the alignment of the eye.
The image of the retina portion as appearing within the iris edge 36E would
usually appear as a dark hole in a bright surrounding image because the
reflectivity of the iris is much greater than that of the retina in most
cases. The substantial detail within the darker retinal image, especially
the visible fine blood vessels near the optical disc are quite suitable
for computer vision processing. For example, processing can be used to
locate and indicate the pupil centroid 46. Various known pattern
recognition techniques may be used to compare retinal detail to a
reference image of the same eye.
The imaging system 12 used for eye alignment must meet relatively demanding
performance requirements to accomplish its full purpose. The eye alignment
information must be obtained and indicated in a time that is short
compared to the time during which the eye normally changes its alignment
significantly. A typical acceptable time is believed to be about 30 to 50
milliseconds. Within a fraction of this time, the imaging system 12 must
form an image with sufficient resolution and depth of focus to provide
accurate measurement of the pupil centroid and the retinal detail.
Typically, the relatively low reflectivity of the retina means that it
must be illuminated artificially to satisfy these requirements, but
injurious light levels must be avoided. Further, it is best to avoid
uncomfortable light levels because eye responses such as accommodation,
aversion or blinking may result and may invalidate a diagnostic
measurement. The imaging system should also provide frequent images in
order to follow eye motion to guide the approach to proper alignment.
By using the general principles of operation of the scanning laser
ophthalmoscope (SLO) co-invented by one of the present inventors and
described in more detail in the incorporated by reference U.S. Pat. No.
4,213,678, one can provide an imaging system which will meet the criteria
specified above. The SLO has two major advantages for providing alignment
information based upon the showing of a dual iris image and retinal image.
In particular, the SLO provides retinal images at TV rates with relatively
low levels of retinal illumination. Further, the SLO, as modified for the
present purposes, may provide high resolution with the large depth of
focus required to image both the pupil and the retina.
FIG. 4 shows the eye alignment imaging system 12 in position for measuring
a subject's eye 10, while the eye 10 is, at the same time subject to
measurement by a Computer Modeling System (CMS) keratograph having an
illumination head 50 and a camera 42. The illumination head 50 and the
camera 52 are part of the CMS manufactured by Computed Anatomy, Inc. and
designed to provide topographic measurements of a cornea. The CMS head 50
and camera 52 are simply shown in FIG. 4 to illustrate how the present
invention may be used in combination with a particular known diagnostic
instrument. Additionally, the head 50 and camera 52 are included in FIG. 4
to show a specific manner in which the present imaging system 12 may be
used without interference between it and another eye diagnostic
instrument. Specifically, the illumination head 50 may use red light and a
filter 54 may only allow red light to pass through to the camera 52. In
that fashion, the camera 52 will avoid errors which might otherwise be
caused by light from the imaging system 12 since the imaging system 12 may
use green light from laser source 56.
The green light from laser light source 56 is supplied to a scanner 26
which may scan the beam 28 in a pattern in order to image a retinal
portion and an iris portion of the eye 10 as discussed in detail
previously. Numerous scanning patterns could be used, but the preferred
scanning pattern resembles a conventional interlaced television scanning
pattern except that image scanning is continued during the backward sweep,
this pattern being discussed in more detail in the incorporated by
reference Pomerantzeff patent. The scanner 26 may include various optical
elements as will be discussed in detail below. The light from the scanner
26 will pass through the beam splitter 58 on its way to the eye 10. The
beam splitter 58 is preferably a polarizing beam splitter. Since the
scanner 26 may supply polarized light (by use of a polarized laser 56) and
the polarized light is about 98% transmitted through the polarizing beam
splitter 58, most of the beam energy from scanner 26 will pass through to
the eye 10. Additionally, the reflections from the cornea and lens of the
eye 10, which are interfering signals for present purposes, will generally
still be polarized such that these reflections will pass through the
polarizing beam splitter 58 and will not be directed towards a light
detector 60. However, unlike the reflections from the cornea and lens, the
scatter returning from the iris and retina are substantially depolarized
such that about 50% of the light returning from the surface of interest
will be deflected towards the light detector 60. Thus, use of a polarizing
laser 56 and a polarized beam splitter 58 allows one to substantially
increase the strength of light returning from the surfaces of interest
relative to reflections from surfaces which are not of interest for
alignment purposes. The reflected light supplied to light detector 60 goes
through filter 62 which would be a green filter corresponding to the green
light from laser 56. The filter 62 will minimize the response of the light
detector 60 to reflections from the red light used by the illumination
head 50.
In addition to having the scan crossing point 34 disposed in front of the
eye as shown, note that the system does not include any optical elements
between that crossing point 34 and the eye 10. Among other modifications
to the SLO disclosed by the Pomerantzeff incorporated by reference patent,
the diameter of beam 28 would generally be significantly smaller than the
diameter used for high resolution at the retina when the SLO is used as a
fundus camera in accord with the Pomerantzeff patent. By use of a smaller
beam than in the fundus camera application of the SLO, one can get a good
focus on the pupil, while at the same time having a sufficient depth of
focus that the retina features can be used for alignment purposes. It
should briefly be noted that the SLO fundus camera as described in the
incorporated by reference Pomerantzeff patent uses a stop 88 in FIG. 1 of
that patent having an aperture therein which corresponds to the size of
the pupil, thereby preventing the detector from detecting light reflected
from the iris. Although the present system 12 may include a stop 64, the
stop will have an aperture which is sufficiently wide to allow light from
the iris to pass there through. In other words, the light detector 60 will
be able to receive reflected light from the iris as well as the retina.
The stop 64 may avoid reflections or stray light from outside the area
upon which the light beam 28 is applied.
The light detector 60 is connected to an amplifier 68 which provides an
output to a scan converter 70. The scan converter 70 in turn supplies an
output to a real-time monitor 72. Additionally, the scan converter 70
might be connected to a print-out device (not shown). Generally, the
operation of the amplifier 68, scan converter 70, and monitor 72 will be
the same as the corresponding components in FIG. 2 of the Pomerantzeff
patent. It should also be noted that, for ease of illustration, the timing
and synchronization, scan amplitude and centering, X driver and Y driver
blocks have not been shown in FIG. 4, it being readily understood that
these may operate in the same manner as shown and discussed relative to
FIG. 2 of the Pomerantzeff patent.
As shown in FIG. 4, the arrangement may also include a storage means 74
which may store the image of a particular eye alignment (i.e., a combined
iris image and retinal image which reveals eye alignment information). The
storage means 74 may be a digital memory or any of various known types.
Alternately, one could use a videotape recorder (not shown) to store past
alignment data, in which case the video tape recorder might be connected
directly to the output of amplifier 68.
Regardless of the type of storage means 74 which is used, it may supply a
signal to a past image monitor 76. The monitor 76 will show a previous eye
alignment in accord with the digital memory, video tape, or other type of
storage means 74. Although FIG. 4 shows two separate monitors 72 and 76,
these monitors might simply be separate halves of a cathode ray tube (not
separately shown) by use of known splitscreen techniques. In that case,
one half of the screen would show the past combined iris/retinal image,
whereas the other half of the screen would show the real-time combined
iris/retinal image. The two separate monitors 72 and 76 may collectively
be considered as a monitor means.
By use of the storage means 74, one can use the system 12 to determine that
the eye 10 is back into a previous orientation. For example, the eye 10
may be in a relatively arbitrary position when the CMS keratograph is
first used. When the keratograph is used, one would also use the imaging
system 12 to provide a combined iris/retinal image which would be stored
within the storage means 74. A doctor or other person may then perform
some therapeutic procedure upon the eye 10. In order to determine what
effect, if any, the therapeutic procedure has had upon the topography of
the cornea, one would like to remeasure the cornea. However, in order to
have meaningful data, the eye 10 must be oriented the same way it was when
the cornea measurements were first made.
Ensuring that the post-therapeutic cornea measurements are made with the
eye in the same orientation as the pre-therapeutic measurements, one would
first position the eye 10 relative to the imaging system 12 to allow it to
image the eye. The imaging system 12 would then provide an iris image of
the iris portion of the subject's eye and a retinal image of a retinal
portion of the eye. This post-therapeutic combined image is then used to
determine objective eye alignment. More specifically, the doctor or other
medical professional could move the subject's head, or, more directly, his
eye until the real-time monitor 72 shows a combined image which is
substantially identical to the combined image of the pre-therapeutic
combined image which may be on display in the past image monitor 76. In
order to return the eye 10 to its same alignment as in the
past-therapeutic combined image, the person manipulating the eye would
ensure that the relative positions of the pupil centroid 46, fovea 42, and
optical disc 44 (refer back momentarily to FIGS. 3A, 3B, and 3C) are the
same as in the pre-therapeutic combined image. To that end, the
positioning of the crossings of blood vessels 48C (again referring to
FIGS. 3A, 3B, and 3C) and the blood vessels themselves may be especially
helpful in establishing that eye alignment is the same for the
post-therapeutic measurement as it was for the pre-therapeutic
measurement. Indeed, one might use several features of the blood vessels
around the optical disc to return to a previous alignment without
necessarily specifically considering the position of the fovea.
In addition to using various known computer vision processing techniques to
automatically supply the centroid 46 (FIGS. 3A, 3B, and 3C), one could
alternately simply insert the centroid in the image by the operator
locating the centroid and inserting it using known technology to insert
points or marks on an image.
Upon the past image monitor 76 and the real-time monitor 72 showing that
the post-therapeutic eye alignment is the same as the pre-therapeutic eye
alignment, one may then turn on the CMS keratograph to provide the cornea
measurements. Since the alignment of the eye 10 is the same as it was when
the original cornea measurements were made, differences between the cornea
measurements before and after the therapeutic procedure can reasonably be
said to result from the therapeutic procedure.
As an alternative to having the operator bring the eye 10 into the same
alignment for the post-therapeutic measurement as the alignment in the
pre-therapeutic measurement, the past combined iris/retinal image may be
compared to the real-time combined image and the difference in orientation
could be calculated. In that case, the cornea measurements may be made in
a post-therapeutic alignment which is different than the pre-therapeutic
alignment, but computer software could be used to provide a transformation
to convert the post-therapeutic cornea values into corresponding
pre-therapeutic values. To take a simple example, if the pre-therapeutic
measurement was made with the pupillary axis in line with the instrument
axis, and the eye alignment imaging system 12 shows that the
post-therapeutic cornea measurements are being made with the pupillary
axis still in line with the instrument axis, but with the eye rotated
30.degree. counter-clockwise about the pupillary axis, the
post-therapeutic cornea measurements would be transformed by, in effect,
rotating the cornea measurements 30.degree. counter-clockwise about the
pupillary axis.
In addition to allowing one to determine eye alignment information by
comparison of before and after combined iris/retinal images, a single
combined iris/retinal image might be used to determine objective eye
alignment in the same fashion as discussed above during the initial
explanation of FIG. 3A. In other words, the relative positioning of the
fovea 42, optical disc 44 and centroid 46 in FIG. 3A was indicative of the
pupillary axis being directed about 7.5.degree. to the left of the
alignment instrument axis. Depending upon the degree of accuracy which is
required in a particular application, one might assume that the pupillary
axis corresponds to the visual axis such that information about the
direction of the pupillary axis could be used to infer what the subject
(such as a test animal) was viewing. Since the visual axis does not pass
through the centroid of the pupil for all subjects, this assumption that
the pupillary axis and visual axis coincide would be an inappropriate
assumption under other circumstances.
Upon determining the eye alignment by the relationship between the
pupillary axis and the instrument axis (and preferably also including some
indication about the rotation of the eye), one could continue to track
this eye alignment data with the alignment imaging system 12 of the
present invention during a therapeutic procedure upon the eye 10.
Alternately, one could obtain this alignment data and, in effect, transfer
the alignment data to an eye tracker. The eye tracker is a known
instrument which is simpler than the present invention, but normally
yields incomplete or inaccurate alignment data. By using one of several
known types of eye trackers to which the alignment data from system 12 has
been transferred, the system 12 may be removed from its position whereas
it images eye 10 such that the system 12 will be out of the way of any
therapeutic instruments which might be used on the person's eye. The eye
tracker will provide information based upon a starting point supplied by
the system 12.
The scanner 26 of FIG. 4 may be constructed in any of various generally
known ways to provide the desired scan pattern as discussed above.
However, with reference now to FIG. 5, an especially advantageous manner
of constructing the scanner 26 will be discussed. The laser light source
56 supplies light to a reflector 78 which reflects the light to a beam
expander 80. An acousto-optic deflector 82 may be used to provide TV rate
deflection without moving parts or excessive sound. However, the
acousto-optic deflector has a large aperture. As the deflector 82 is a
diffraction limited device, the number of resolvable positions is
proportional to the aperture in the direction of scan. In order to supply
a sufficiently large number of resolvable positions, acousto-optic
deflectors have large apertures only in the necessary directi | | |
