The laser engine used to perform the ablations was a Q-switched,
flashlamp pumped Nd:YAG laser (Surelite II; Continuum, Santa Clara, CA)
capable of producing up to 660 mJ of energy per pulse at a
wavelength of 1064 nm. The duration of the 1064-nm pulses was 5 nsec
with a repetition rate of 10 Hz. To produce the fifth harmonic
wavelength, three nonlinear crystals were used. The second harmonic
(532 nm) was produced using the crystal BBO (Casix Inc., Fuzhou,
China). The fourth (266 nm) and fifth harmonics (213 nm) were produced
using CLBO crystals (Crystal Associates, Waldwick, New Jersey).
The conversion efficiencies obtained were 60%, 25%, and 40%
respectively, for each stage, resulting in an overall conversion
efficiency of 6%. The maximum fifth harmonic output energy was 20 mJ
per pulse for a fundamental input energy of 330 mJ.
Ablations were carried out on porcine corneas obtained from an
abattoir. All eyes were used within 5 hours of enucleation and stored
on ice until required for ablation. Immediately before laser
irradiation the epithelium was débrided and the eye vacuum
stabilized onto a clamped syringe. The eye was positioned less than 5
mm behind a 0.5 × 2.5-mm slit onto which the collimated laser
beam was centered. Therefore, the slit acted as an aperture for the
beam rather than using focusing to produce the required beam diameter.
The dimensions of the slit were chosen to produce a rectangular
ablation profile to simplify the task of measuring the depth. The pulse
energy through the 1.25-mm
2 slit was measured before and
after each ablation using a power meter (model 1825-C; Newport, Santa
Ana, CA) with a detector (model 818T-10; Newport). A small fan was used
for each ablation to disperse the ablation plume to minimize any
absorbing or blocking effect this would have on the impinging laser
radiation.
12 13
For each ablation trial, the fluence was held constant, and a number of
eyes were ablated with a varying number of pulses. This allowed the
ablation rate, defined as the etch depth per pulse to be obtained for a
wide range of fluence values. This procedure was repeated for fluence
values of 80, 120, 150, 180, 210, and 250 mJ/cm2. The
number of pulses was varied between 100 and 400. A separate trial was
undertaken to determine the fluence value for the threshold of
ablation. The method used for this trial was to irradiate eyes with low
values of fluence and a large number of pulses until a notable etching
effect was observed from a profiling system (described later).
In an effort to eliminate potential depth measurement errors associated
with histologic methods,
14 15 the ablated eyes were
analyzed immediately after ablation. Surface profiles of the ablated
corneas across the 0.5-mm cut were measured using a laboratory
custom-made, computer-automated confocal profiling system. The design
of the profiling system is similar to a confocal microscope and has a
depth resolution of 10 μm. Although the resolution is greater than
the submicrometer ablation of each pulse, it is suitable for measuring
the depth of an ablation resulting from a large number of pulses. Also,
a resolution of this magnitude provided a convenient working distance
and facilitated surface detection. Calibration of the profiler was
achieved using an interferometric method. This involved using the
actuator that was used to oscillate the microscope objective to
oscillate a plane mirror in one arm of an interferometer. This allowed
accurate calibration of the actuator position. Immediately after
irradiation, the eye was placed in the profiler while affixed to the
syringe to maintain intraocular pressure. The eye and syringe were
placed on an
x–
y translation stage
(Newport) and positioned in the beam of the profiler so that it was at
the latitudinal center of the cut. The microscope objective used to
focus the beam onto the surface of the cornea was oscillated with a
frequency of 25 Hz. This enabled the depth measurement at each position
across the cut to be averaged over many measurements to minimize error.
The profile data were then analyzed to obtain the depth of ablation.
The corneal profile data obtained from the confocal microscope were
separated into two data series. The first represented the unablated
corneal profile. The second series represented the ablated segment of
the cornea. Additional data series were then derived and used in the
calculation of ablation depth.
Figure 1 shows one such profile. To obtain a meaningful measurement of the
ablation depth, the need arose to extrapolate the unablated curvature
of the corneal surface. In this way, the depth could be calculated from
the surface extrapolation rather than from a straight line across each
edge of the cut. This extrapolated curve was determined by applying a
second-order polynomial fit to the corneal surface data points. To
calculate the depth of ablation, another two polynomials were used.
Each of these polynomials was parallel to the fitted curve of the
unablated corneal surface. The first was drawn through the deepest
point of the ablated region. The second is termed an acquisition line
and was heuristically set at 75% of the distance between the surface
fit and the deepest point of the ablation. The acquisition line
determines which data points from the ablated region are included or
acquired for calculation of the cut depth. Included points are
therefore those located between the deepest point and the acquisition
lines. The magnitude of the depth of each included point was taken as
the vertical distance from the corneal surface fit. Those data points
that are included are then averaged and an ablation depth obtained. A
line displaying the averaged ablation depth is located in the central
region of the cut in
Figure 1 .