Development of photorefraction techniques for studies in humans
and various animals. Development of a video gaze tracker. Optics
of the eye. Models and analysis of the mechanisms of myopia, based
on the experimental results from the laboratory. Pattern electroretinograms
in chickens.
Photorefraction
in humans
The principle of eccentric photorefraction is
as follows:
Light emerging from a light source (in this case, a high power infrared
LED) in the camera aperture creates a bright spot in the back of
the eye.
The spot, in turn, serves as a light source itself. It is imaged
back in space and its best focus is in the plane of where the eye
is focussed. If the eye is focussed into the plane of the video
camera of the PowerRefractor (A), the pupil is about homogeneously
illuminated with no gradient in the brightness distribution (B -
top) since the rays return to the initial light source, the LED.
However, if the eye is focussed close (myopia,
B - middle), the rays diverge behind the plane of focus. If now
a part of the camera aperture is covered by a shield (black line
in B - middle and bottom), rays returning from the top of the pupil
cannot enter the camera aperture.
As a result, the pupil is illuminated only in its lower part. Conversely,
if the eye is hyperopic (B - bottom), only the rays emerging from
the top of the pupil enter the camera aperture and the pupil is
illuminated in the top.
Fig. 2: Appearance of the brightness distribution
in the pupil in myopic eyes
If several rows of infrared LEDs are used the
brightness distribution in the pupil can be fit by linear regression
with high correlation coefficients. The nice thing is that the slope
of the regression is linearly related to the refractive error of
the eye in the meridian perpendicular to the knife edge of the photoretinoscope.
After calibration with trial lenses, the subject's refractive error
can be sampled with high resolution.
Automated photorefraction is particularly powerful
in measuring the dynamics of accommodation binocularly (yellow and
blue traces on the right). Sampling occurs at 25 Hz (European Video
format). Pupil sizes (traces on the left) and the convergence of
the pupil axes of the eyes (red trace) are also recorded. In this
measurement, the subject wears a –2 D trial lens in front of
its left eye. Therefore, refractions are offset between both eyes
by 2 D (yellow and blue traces on the right). Accommodation amplitude
was about 2 D.
Fig. 3
For a video showing accomodation measurements
see acc.mpg
acc.mpg,
726 KB
Our work on photorefraction merged into a commercial
machine, the PowerRefractor. The PowerRefractor was tested in patients
in the Eye Hospital, in the Kindergarten, and in the laboratory
(Choi et al,
2000). Furthermore, it is under testing in a number of laboratories
around the world.
The PowerRefractor, including its patent, was completely taken over
in August 2001 by "Plusoptix"
in Nürnberg, Germany. Since 2003 the company is developing
other versions of the original refractor. Frank Schaeffel is not
involved in these developments and is not responsible for their
performance.
About 50 publications can be found in PubMed, using the PowerRefractor.
Photorefraction was applied in a wide variety of animals, including:
Mice appear to have poor optical quality in
their eyes, as judged from the inhomogeneous illumination patterns
in the pupil (compare to humans, below).
However, since the ring shaped brightness patterns are consistent
among individuals, it could also be that they indicate the presence
of variations in focal length across the pupil which may have a
"deeper sense", similar to results from fish eyes, described
by Ronald
Kröger and co-workers.
Aberrations of the mouse eye were described in detail in a collaborative project with Elena de la Cera and Susana Marcos, Madrid.
de la Cera EG, Rodríguez G, Llorente L, Schaeffel F, Marcos S. Optical aberrations in the mouse eye Vision Res. 2006 Aug;46(16):2546-53
Fig. 4
Fig. 5: Mouse refraction
For a video showing dynamic photorefraction
in a normal and a myopic mouse, see mouseref.mpg
and mouseDM.mpg.
mouseref.mpg,
751 KB
mouseDM.mpg,
1.206 KB
This technique also permits highly precise and convenient pupillography. The new generation USB2 IR cameras grab frames at 60 hz and software was developed to permit pupillography on any labtop, using the USB port also to control a green stimulation LED.
Fig. 6: Appearance of the labtop screen during pupillography in alert mice.
Using a highly magnified video image of the
pupil and sub-pixel resolution programming, the pupil centre and
the first Purkinje image (see red arrow) were localized at 25 Hz
so that gaze tracking was possible with about 0.2 deg angular resolution.
The technique works from 90 cm distance so that the subject's eye
position can be tracked while working on a computer.
Fig. 7: Gaze measurement during reading
Fig. 8: Gaze measurement with 25
hz (2001), with random fixation point
Automated optomotor experiment to measure spatial vision in mice
Different from other published mouse optomotor response measurement devices, the mouse could freely move in the inner drum (see videos: optomotor1.mpg and optomotor2.mpg). Its angular movement of the centre of mass, as well as the body orientation were automatically tracked and statistically analyzed.
The technique was published first in Christine Schmucker's paper in Investigative Ophthalmology and Visual Science in 2005 (--> IOVS).
Measurementsof lens tilt and decentration in normal and pseudophakic subjects
Inspired and supported by Dr. Hakan Kaymak and Prof. Ulrich Mester, Sulzbach, Germany, a portable device was constructed and programmed that measures lens tilt and decentration from a distance of about 25 cm (Fig. 10).
A PowerPoint presentation gives details about the procedure.
The study is currently in press at Investigative Ophthalmology and Visual Science.
