able optical
Ot represent
Any one
1988) have
points for
in aerial
applied in
s and Petsa
st fit surface
] by Pilgrim
994, 1995).
ion between
ninimise the
d Koivunen
an may be
ind corneal
que that is
opriate than
le methods
he utility of
the surface
st fit radius
e of corneal
n a limited
ower and to
er of points
cy required,
rs have not
ie following
to compute
cap along a
ute reliably
a over any
limensional
curacy and
instruments
g spherical
it that uses
> the radius
itely than it
e accuracy
. Keratocon
rithms used
ver and a
by current
y addressed
accuracy in
procedures
94).
6. CONCLUSIONS
An accurate knowledge of the topography of the anterior
surface of the human cornea is essential in a wide range of
ophthalmic applications. None of the commercially available
methods of measuring corneal topography provide an
accurate topographic map of the entire cornea, particularly
when the cornea is diseased or scarred.
A photogrammetric method of measuring corneal topography
has been developed. The method does not rely on corneal
reflectance and no assumptions need to be made about the
approximate or specific geometry of the cornea. Using non-
metric cameras, an accuracy of +20um has been obtained.
The instrument has provided complete and accurate
measurements of corneas not well suited to current
commercially available corneal mapping systems, particularly
grafted and keratoconic corneas.
Full automation of image matching is likely to be the greatest
problem to be overcome in a digital implementation of the
Keratocon. The task is to obtain sufficient contrast between
the targets and the pupil, iris and sclera to allow
segmentation of the targets from the background. Additional
experimentation with a teflon membrane appears to be the
most likely route to a reliable solution.
The ophthalmic market place has demonstrated that many
clinicians will be satisfied with an instrument that only maps
corneas that have relatively small departures from normality.
Given these conditions, and the fact that a photogrammetric
instrument is unlikely to be less expensive than a
videokeratoscope, it appears unlikely that a digital
implementation of the Keratocon is commercially viable if its
only advantage is that it provides higher three dimensional
accuracy because of its rigorous photogrammetric solution.
The key to producing a commercially viable instrument is
incorporating into the instrument's software the capacity to
apply the three dimensional data to high quality data analysis
and visualisation. Existing corneal mapping systems still
provide only crude visualisation and analysis tools. A digital
Keratocon should provide sophisticated visualisation of
corneal topography, be able to derive parameters such as
corneal power, and have surface matching capability.
Medical photogrammetry, if it is to be of use to the
practitioner, must extend beyond the task of measurement
alone. Spatial measurement in isolation is generally of
limited value. The task is more likely to be an analysis of
temporal change, long term in the case of progressive
disease, short term in the case of motion analysis; or a
measurement that must be linked to function, such as the
functional link between corneal topography and refractive
power.
449
REFERENCES
Abdel-Aziz YI (1974) Expected Accuracy of Convergent
Photos. Photogrammetric Engineering 1974: 1341-
1346.
Abdel Aziz YI, Karara HM (1971) Direct Linear
Transformation from Comparator Coordinates into
Object Space Coordinates in Close-Range
Photogrammetry. Proceedings of the Symposium on
Close-Range Photogrammetry, Urbana 1971: 1-8.
Banda FAS, Muller J-P (1990) Cornea Shape Measurement.
SPIE Vol.1395 Close-Range Photogrammetry Meets
Machine Vision 1992: 720-726.
Bogan SJ, Waring GO, Ibrahim I, Drews C, Curtin L (1990)
Classification of Normal Corneal Topography Based on
Computer-Assisted Videokeratography. Arch Ophthal
Vol.108 1990:945-949.
Bonnet R (1959) Stéréophotogrammétrie de la Cornée
Humaine. Revue d'Optique, T.38, Nos. 9, 10, 1959:
447-460.
Bonnet R, Cochet P (1962) New Method of Topographical
Ophthalmometry — Its Theoretical and Clinical
Applications. American Journal of Optometry and
Archives of American Academy of Optometry Vol.39
No.5 May 1962: 227-251 translated from Bulletins et
Mémoires De La Société Francaise D'Ophthalmologie
Vol.73 1960: 688-716.
Bores LD (1991) Corneal Topography — The Dark Side of
the Moon. SPIE Vol.1429 Holography, Interferometry,
and Optical Pattern Recognition in Biomedicine 1991:
217-228.
Clark BAJ (1973a) Conventional Keratoscopy — A Critical
Review. Aust.J.Optom. Vol.56, 1973: 140-155.
Clark BAJ (1973b) Variations in Corneal Topography.
Aust.J.Optom. Vol.56 1973: 399-413.
Clark BAJ (1974) Mean Topography of Normal Corneas.
Aust.J.Optom. Vol.57, 1974: 107-114.
Dingeldein SA, Klyce SD (1989) The Topography of Normal
Corneas. Arch Ophthalmol 107 1989:512-518.
Duke-Elder S (1970) System of Ophthalmology Vol.5:
Ophthalmic Optics — Refraction. CV Mosby, St Louis,
1970.
El Hage SG (1972) Photokeratoscopy by Using the Cornea as
a ‘Self Luminance System. Am.J.Opt. & Arch.
Am.Ac.Optom. July 1972: 595-602.
El Hage SG (1976) The Three-Dimensional Configuration of
the Cornea. Nouv.Rev.Optique, t7, No.3 1976: 205-207.
El-Hakim SF, Moniwa H, Faig W (1979) Analytical and
Semi-Analytical Photogrammetric Programs — Theory
and Users Manual. Technical Report No.64 Department
of Surveying and Engineering, University of New
Brunswick.
Friedlander MH, Mulet M, Buzard K, Granet N (1991)
Holographic Interferometry of the Corneal Surface.
Proceedings of SPIE Vol.1429 Holography,
Interferometry and Optical Pattern Recognition 1991:
229-236.
Gormley DJ, Gersten M, Koplin RS, Lubkin V (1988)
Corneal Modelling. Cornea 7(1) 1988:30-35.
Guillon M, Lydon DPM, Wilson C (1986) Corneal
Topography: A Clinical Model. Ophthalmic
Physiological Optics, Vol.6 No.1 1986: 47-56.
Heath GG, Wheeler WH, Soni PS, Horner DG (1991)
Reliability and Validity of ^ Videokeratoscopic
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996
