Download Light scattering from the human cornea.

Light scattering from the human cornea
Thomas Olsen
A clinical method of measuring angular scatter from the human cornea is described. The
method, entails the use of a slit-lamp photometer with a pin-light attachment to control the
angle of measurement. Corneal scatter and corneal thickness were measured in 93 normal
subjects and 56 patients recently operated on for cataract. In the latter group, the surgically
induced, increase in corneal thickness was found, to increase corneal scatter markedly, demonstrating the well-known dependence of scatter on corneal hydration. For the normal cornea,
thickness and. scatter were only insignificantly correlated. This finding was interpreted as
evidence that the normal variation in corneal thickness is caused by a variation in mass content
rather than in water content of the cornea. An increase in corneal scatter with age was
demonstrated. Because this was seen at a constant corneal thickness, the mechanism was
presumably that of age-related alterations in the latticelike organization of the collagen fibrils.
These results suggest that clinical measurements of scatter may be used as a means to study
hydration and ultrastructural characteristics of the human cornea in vivo. (INVEST OPHTHALMOL Vis Sci 23:81-86, 1982.)
Key words: cornea, light scatter, thickness, hydration, photometry, in vivo
o
ptical clarity is one of the fundamental
properties of the cornea. Although the scattering properties of the cornea have been the
subject of several theoretical and experimental in vitro studies,1-l2 no quantitative method
seems yet to have been described that has
been applied to living human cornea. The
purpose of the present study was to describe
such a method and to present data on the
normal variation in corneal scatter and its dependence on corneal thickness.
Materials and methods
Apparatus. The light detection unit (Fig. 1) was
basically identical to that described by Waltman
From the Department of Ophthalmology, Aarhus
Kommunehospital, University of Aarhus, Aarhus,
Denmark.
Supported by the Danish Committee for Prevention of
Blindness and Landsforeningen for Sukkersyge.
Submitted for publication July 29, 1981.
Reprint requests: Dr. Thomas Olsen, Department of
Ophthalmology, University of Aarhus, Aarhus Kommunehospital, DK-8000 Aarhus C, Denmark.
and Kaufman13 for slit-lamp fluorophotometry.
This method employs a fiber optic probe incorporated into the eyepiece, whereby light is picked
up in a small area of the image plane and guided
into a photomultiplier. To increase the stability of
the present system, a synchronous "lock-in" detector (Princeton Applied Research Corp., Princeton, N. J.) was coupled to the current amplifier of
the photomultiplier and the slit light was chopped
at about 400 Hz with a rotor blade (Rofin Inc.,
New Upper Falls, N4ass.) within the lamp house
(of a Zeiss slit lamp). The output was recorded on
an X-Y recorder that could be activated by a foot
switch.
The light source was the slit-lamp tungsten
bulb. Blue light was used throughout the present
investigations by inserting an interference filter
that passed wavelengths of 400 to 500 nm (made
by OPLAB, Lyngby, Denmark). The width of the
focused slit beam was about 160 /xm and could
be reproduced with the aid of a stop on the slit
aperture.
Preliminary experiments were conducted by
asking the subject to look into the slit light while
the observer viewed the optical section of the central cornea at a fixed angle to the incident slit light.
0146-0404/82/070081+06300.60/0 © 1982 Assoc. for Res. in Vis. and Ophthal., Inc.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933104/ on 05/10/2017
81
Invest. Ophthalmol. Vis. Sci.
July 1982
82 Olsen
Fig. 1. Diagram of the instrument used for the angular measurement of corneal scatter.
However, difficulties associated with a correct
alignment of the incident angle on the cornea
made these measurements vary to a large extent.
To control the geometric configuration of the measuring situation further, two pin lights were
mounted on the slit lamp as shown in Fig. 1. A
correct alignment for a reading to be taken could
then be defined as that position of the apparatus
where the pin lights were seen through the eyepiece to be reflected at the anterior limit of the
optical section of the cornea (Fig. 2) while the patient gazed into the slit light. This principle was
identical to the modification described by Mishima
and Hedbys14 for measurements of corneal thickness. The use of a chopped signal made the instrument insensitive to the pin lights.
Sample geometry. In the image plane of the microscope, light was sampled in an area correspond-
ing to the cut end of the fiber optic probe. The
measurements were made with the image of
the optical section of the cornea falling through the
center of the sample area. The diameter of the
probe was 450 fxm. The magnification of the objective lens was 0.48X, giving a projected sample
diameter of 938 jxim in the object plane. An angle
of 45° was chosen between incident slit light and
microscope. Viewed at this angle, the width of the
optical section of the cornea is about half the
thickness of the cornea, i.e., 0.25 mm for a 0.50
mm cornea or 25% of the diameter of the sample
area. (The exact values can be calculated as described by Olsen et al.15) To obtain a linear response between thickness and scatter, a rectangular rather than a circular shape of the sample area
would be optimal. Calculations showed, however,
that the deviation from a linear response amounted
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933104/ on 05/10/2017
Volume 23
Number 1
Light scattering from human cornea 83
Table I. Age dependence of normal corneal
scatter and thickness (mean ± S.D.)
Age group
Cornea! scatter
(arbitrary units)
Corneal thickness
(mm)
20-40 yr
(n ~ 36)
60-80 yr
(n = 39)
0.609 ± 0.061
0.793A ± 0.127B
0.545 ± 0.031
0.547 c ± 0.035c
A
Significantly different (p < 0.001) from younger group by t test.
Significantly different (p < 0.001) from younger group by F test.
Not significantly different (p > 0.05) from younger group.
c
to less than 4% for corneal thicknesses up to 1 mm,
making a correction unnecessary.
The angle of 45° was sufficiently large to avoid
lens scatter entering the sample area for anterior
chamber depths greater than 1 mm.
Procedure and standardization. The subject
was seated in front of the slit lamp and was asked
to gaze directly into the slit light. When a correct
alignment had been achieved as described above,
a reading of about 2 sec duration was taken. The
reading was repeated twice after intervals for the
subject to blink. An average of these triple readings was taken as "one" measurement. Immediately after completion of the measurement, a reading was taken of a 10~6 gin/ml sodium fluorescein
solution. The fluorescein standard was kept in a
dark place and renewed every month. All measurements were expressed relative to this standard,
which was given the value 1.0. The reading of the
standard was found to be equivalent to about 50
international units of opacity (World Health Organization Fifth International Reference Preparation
of Opacity, 1976, obtainable from National Institute for Biological Standards and Controls, Hampstead, London, England). The latter standard,
however, was not well suited for quantitative measurements of angular scatter and was impractical as
a routine standard.
Subjects, The normal corneas were from students or members of the medical staff or from patients admitted with no pathologic ocular conditions other than senile lens opacities. Patients with
corneal edema had been operated on for cataract
less than 1 week before testing. Patients with epithelial edema, increased intraocular pressure, or
anterior uveitis were excluded from the present
series.
Corneal thickness was measured with a modified Haag-Streit pachymeter. Single determina-
Fig. 2. Composite photograph showing position of
the optic probe (white circle) and reflections of the
pin lights (stars) in the optical section of the cornea.
tions using alignment method "A" and a slit width
of 30 /Ltin were done according to the methods of
Olsen et al., IS and the readings were corrected for
nonlinearity.15 Both the thickness and the scatter
measurements were taken in the afternoon hours.
Statistical analysis. Unless specified otherwise,
conventional distributional methods were used.
Linear regression analyses were performed with
the method of least squares. 17
Results
The instrument was easy to handle and
caused no discomfort to the subject. From
repeated measurements on more than 20
subjects, the day-to-day variation was found
to be about 7%, expressed as the coefficient
of variation (S.D./x). The intrasession variation amounted to about 4%. These variations were fairly constant over the measuring
range. The measurements were insensitive to
variations in slit height (as long as the entire
sample area was covered), and the average
values for light and dark irides did not differ.
This indicates that stray light did not influence the readings.
For normal corneas, a significant correlation was found between the scatter and the
age of the subject (Fig. 3). Furthermore, the
interindividual variation in scatter increased
with age (Table I). Corneal thickness did not
parallel these age-related changes.
Because of the dependence of scatter on
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933104/ on 05/10/2017
Invest. Ophthalmol. Vis. Sci.
July 1982
84 Olsen
Scatter
Scatter
1.0-
1.0-
fv
0.5-
10
20
30
40
50
t
0.5-
60
70
80
90
•?
H
0.50
0.55
0.60 MM
Fig. 3. Corneal scatter vs. age in 93 normal subjects. Spearman s rank correlation coefficient rs =
0.55 (p < 0.001).
Fig. 4. Corneal scatter vs. thickness for a normal
group 20 to 40 yr of age.
age, the analysis of the influence of corneal
thickness on scatter was divided into a young
and an old group (Figs. 4 and 5). In each
group, corneal thickness and scatter could be
reasonably approximated by a normal distribution. No significant correlation was observed between corneal scatter and thickness
in either group. Linear regression analysis (y
on x) of these data showed the regression
lines to have an intercept on the y-axis, which
was significantly different from zero (Table
II). No systematic difference was found between readings from right and left eyes or
between sexes.
The influence of corneal edema on scatter
is shown in Fig. 6. Most of these patients
were in the age group of 60 to 80 yr, where
normal values can be found in Table I. Although there was some variation, the amount
of light scattered was seen to rise sharply as
the tissue swelled. The curve fitted by the
eye showed an increase in scatter of 0.4 units
for an increase in thickness of 0.1 mm and 1.2
units for an increase of 0.2 mm.
Discussion
The present approach employs measurements of angular scatter from all the layers of
the cornea, which means that scatter from
the limiting layers as well as the stroma were
included in the readings. Attempts at measuring stromal scatter in a more specific way
by using a larger magnification and guiding
the optic probe to a midstroma position of the
optical section met with difficulties associated
with the small saccadic movements of the
eye. A loss of reproducibility was the cost of
this increase in specificity. In the rabbit,
stromal scatter has been found to constitute
more than 75% of total angular scatter in the
backward direction,12 suggesting that measurements of whole corneal scatter are
largely representative of stromal scatter.
The strong dependence of light scatter on
corneal hydration found in the present study
agrees with the results of others. As can be
noted from Fig. 6, the present technique was
capable of detecting increments in scatter induced by slight increases in corneal hydra-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933104/ on 05/10/2017
Volume 23
Number 1
Light scattering from human cornea
85
Scatter
1.0-
•
r
0.50.2
H
0.3
MM
Fig. 6. Increase in scatter in 56 patients recently
operated on for cataract, plotted against the increase in corneal thickness (operated eye minus
unoperated eye).
0.50
0.55
0.60 MM
Fig. 5. Corneal scatter vs. thickness for a normal
group 60 to 80 yr of age.
Table II. Linear regression of the data
shown in Figs. 4 and 5*
Parameter
tion. These results suggest that the method
may be used as an alternative to a corneal
thickness reading in clinical situations where
information regarding the corneal hydration
is wanted.
In the normal range, the corneal thickness
was found to have an insignificant influence
on scatter. Unlike the situation after cataract
extraction, where changes in corneal thickness must be ascribed to changes in corneal
hydration, it is not known what causes the
normal variation in corneal thickness. A linear relationship has been demonstrated between thickness and hydration.18 If the hydration of the cornea was the major factor,
one should expect the scatter to be markedly
influenced by the thickness. However, the
statistical analysis of the data indicated an
upper 95% confidence limit of 0.8 and 1.2
units of scatter/mm for the slopes of the regression lines in the young and the old group,
respectively. This means that the thickness
dependence of scatter in all probability was
Estimated
value
Variance
For the group 20-40 yr (n = 36)
a
0.502
0.0280
b
0.197
0.0941
For the group 60-80 yr (n = 39)
a
0.770
0.1068
b
-0.042
0.3553
95% Confidence
limits f
0.161-0.842
-0.426-0.821
0.108-1.432
-1.249-1.166
*Analysis is made of the parameters a and b of the line y = a +
bx, where y denotes corneal scatter (arbitrary units) and x denotes corneal thickness (mm).
tConfidence limits: estimated value ± to.o5 ' Vvariance, degrees of freedom df = n — 2.
less than 0.08 units of scatter/0.1 mm (about
the normal interval) of thickness in the young
group and 0.12 in the old group. This dependence is much smaller than the increase of
0.4 units of scatter associated with the 0.1
mm increase in water content observed in
the cataract patients. The insignificant correlation with scatter of the normal corneas may
therefore be taken as evidence that the normal variation in corneal thickness is caused
by a variation in mass content rather than in
water content.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933104/ on 05/10/2017
Invest. Ophthalmol. Vis. Sci.
July 1982
86 Olsen
The regression lines of scatter vs. normal
thickness were found to have an intercept on
the ordinate significantly different from zero,
indicating that the relationship was nonproportionate. A possible explanation of this effect may be the limiting layers causing a constant amount of scatter to be included in the
measurements, which did not vary with the
thickness of the cornea.
Corneal scatter was found to increase with
age. In physical terms this suggests the spacing of the collagen fibrils to be less regular
with age, thereby increasing the amount of
light scattered off the forward direction. Because the corneal thickness was observed to
be unaffected in the same age interval, it
seems unlikely that the increase in scatter
was brought about by an increase in corneal
hydration. However, it may be speculated
that other factors such as age-related, degenerative, tear, and repair processes may cause
a primary derangement of the latticelike organization of the collagen fibrils, thereby
causing the cornea to deviate further from its
near transparency.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
REFERENCES
1. Maurice DM: The structure and transparency of the
cornea. J Physiol 136:263, 1957.
2. Kikkawa Y: Light scattering studies of the rabbit
cornea. Jpn J Physiol 10:292, 1960.
3. Kinoshita A, Manabe R, and Kikkawa Y: Angular
distribution of the light scattered from the clouded
cornea. Jpn J Ophthalmol 9:42, 1962.
4. Goldman JN, Benedek GB, Dohlman CH, and
Kravitt B: Structural alterations affecting transpar-
16.
17.
18.
ency in swollen human corneas. INVEST OPHTHALMOL 7:501, 1968.
Hart RW and Farrel RA: Light scattering in the cornea. J Opt Soc Am 59:766, 1969.
Cox JL, Farrel RA, Hart RW, and Langham ME:
The transparency of the mammalian cornea. J Physiol 210:601, 1970.
Feuk T: On the transparency of the stroma in the
mammalian cornea. IEEE Trans Biomed Eng 17:
186, 1970.
Feuk T: The wavelength dependence of scattered
light intensity in rabbit corneas. IEEE Trans Biomed Eng 18:92, 1971.
Benedek GB: Theory of transparency of the eye.
Appl Opt 10:459, 1971.
Lindstrom JI, Feuk T, and Tengroth B: The distribution of light scattered from the rabbit cornea. Acta
Ophthalmol 51:656, 1973.
Twersky V: Transparency of pair-correlated, random
distributions of small scatterers, with applications to
the cornea. J Opt Soc Am 65:524, 1975.
McCally RL and Farrel RA: The depth dependence
of light scattering from the normal rabbit cornea.
Exp Eye Res 23:69, 1976.
Waltman SR and Kaufman HE: A new objective slit
lamp fluorophotometer. INVEST OPHTHALMOL 9:247,
1970.
Mishima S and Hedbys BO: Measurement of corneal thickness with the Haag-Streit pachometer.
Arch Ophthalmol 80:710, 1968.
Olsen T, Nielsen CB, and Ehlers N: On the optical
measurement of corneal thickness. I. Optical principle and sources of error. Acta Ophthalmol 58:760,
1980.
Olsen T, Nielsen CB, and Ehlers N: On the optical
measurement of corneal thickness. II. The measuring conditions and sources of error. Acta Ophthalmol 58:975, 1980.
Armitage P: Statistical methods in medical research.
Oxford, 1977, Blackwell Scientific Publications, Ltd.
Mishima S: Corneal thickness. Surv Ophthalmol
13:57, 1968.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933104/ on 05/10/2017