Photorefractive keratectomy (PRIQ is a procedure that uses an argon fluoride excimer laser (193 nm) to remove a microscopic layer of anterior stromal tissue from the cornea to alter its refractive power. There is virtually no damage to adjacent unexposed tissues.12 The amount and shape of tissue removal required to correct myopic refractive errors has been described previously.3 The photoablation of tissue has been studied, and the effects of hydration and fluence on ablated surface smoothness4"6 and ablation rates79 have been determined. However, the spatial variance of tissue ablation for large diameter areas of cornea exposed to a uniform laser irradiance has not been studied.
Numerous investigators10"15 have demonstrated satisfactory results for patients with low to moderate myopia (less than 6.00 D). The results reported by Kim and associates14 are representative of many other studies, in which almost 90% of eyes had uncorrected visual acuity of 20/25 or better at 2 years following PRK, The average manifest refraction at 2 years (-0.48 ± 0.88 D) was essentially the same as that obtained at 1 year (-0.45 ± 0.88 D). Regression of effect, however, has been found to be more common in patients with high myopia, and the final outcome is thus less predictable for these patients.11,16
Corneal topographic analysis using videokeratography provides objective measures of the quality of the healed anterior corneal surface following PRK. Excimer laser ablations are usually characterized by a relatively uniform distribution of surface power within the treated zone17 and have a smooth power transition to the peripheral cornea.18 In some instances, a central area of undercorrection called a "steep central island" can be demonstrated within the ablated area.19"23 First reported in 1992, the term steep central island is generally defined as a central area of ablation that appears to be flattened less than the surrounding ablated area. Most steep central islands eventually regress, but their presence may cause decreased visual acuity and image ghosting.19,22 In contradistinction to steep central islands, central overcorrection and peripheral undercorrection within the ablation zone have been reported.24"26 This distortion has been associated with the use of small ablation diameters using a nonuniform Gaussian beam profile25 which is hotter centrally.27
Figure 1: Experimental layout showing the optical profilometer, eye, and laser beam.
Many clinical investigators have studied corneal topography following PRK and observed wide variations in quality of corneal topography and the incidence of steep central islands. For example, Hersh et al25 reported no steep central islands while Krueger and associates22 have shown that steep central islands may be seen in over 50% of patients at 1 month following PRK Krueger et al22 defined steep central islands as 3.00 D by 1.5 mm in diameter. In this study, steep central islands were associated with loss of spectacle-corrected visual acuity and image ghosting.
Numerous attempts have been made to determine the extent to which central under-ablation (less than the expected ablation rate) causes steep central islands. For example, corneal topography utilizing videokeratography can be performed immediately after PRK when artificial tears have been instilled.23 However, it is the shape of the tear film and not the ablated cornea which is measured with this technique. A direct measurement of the ablated surface is necessary to truly determine the extent to which central under-ablation causes steep central islands.
Our intent was to determine the shape of the ablated anterior corneal surface immediately after PRK using an optical profilometer. We used this instrument in a series of enucleated porcine eyes to characterize the profile of ablated corneas, to determine the spatial variance of tissue ablation rates, to elucidate factors likely to be associated with steep central island formation, and to evaluate the feasibility of eliminating steep central islands by changing the distribution of laser pulses during PRK. Because changes in ablation geometry could be introduced due to corneal edema and the deteriorating condition of the non-living eyes, we also measured additional ablations in a living animal model.
MATERIALS AND METHODS
Profilometer- The optical profilometer is shown in Figure 1. Light passing through an aperture was collimated by the collimation lens and projected obliquely over the cornea. A second lens imaged the cornea onto a CCD camera (Cohu 4910, Cohu Inc., San Diego, Calif). An aperture closed to approximately 1 mm diameter was positioned at the focal length of the imaging lens to increase the depth of field of the corneal image. This method produces high contrast silhouette images of the cornea. Video images were digitized with a frame grabber (EPIX SVIP, EPIX Inc., Northbrook, IL), and processed with an edge detection program which determined the corneal profile. Subtracting the preablation surface from the post-ablation surface produced the ablation profile. Because slight amounts of motion can produce errors in the ablation measured, the corneal profiles were aligned using unablated portions of each profile prior to subtracting the profiles.
Prior to performing the ablation experiments, we validated the profilometer with spherical and bicurve polymethylmethacrylate (PMMA) surfaces (Con-Cise Lens Co, Fremont, Calif) as a model for PRK. The surface curvature and central ablation zone size were confirmed with a rectilinear photokeratoscope28 which can measure abrupt changes in curvature which would appear smeared on a conventional videokeratograph. A 43.00 D spherical surface served as a model of the preoperative surface. Postoperative surfaces of -4.00 and -8.00 D ablations were modeled using bicurve surfaces with a 43.00 D peripheral surface and 5.5 mm central surfaces of 39.00 and 35.00 D, respectively. A single video frame from the pre-ablation surface and a single frame from the post-ablation surface were analyzed with the edge detection algorithm.
To calculate the ablation shape, the difference between pre-ablation and post-ablation elevation was calculated after the two profiles were aligned. The elevations were determined with an edge detection algorithm which processed each digital image. The elevations were aligned using the peripheral unablated area. This alignment was necessary to prevent errors caused by motion of the enucleated porcine eyes. Prior to and after ablation, the instrument was validated with the above test surfaces.
Excimer laser system- A VISX Twenty Twenty B excimer laser system with Vision Key System and software version 4.10 was used for all ablations CVISX Inc., Santa Clara, Calif). This system produces an output of 193 nm, operates at a frequency of 6 Hz and is adjusted to deliver a uniform fluence of 160 mJ/cm2 with a 6-mm diameter ablation zone. An aspirator positioned 1 3/4 inches above the ablation captures the plume as it lifts from the surface. The PRK ablation algorithm was calculated from the equations of Munnerlyn et al3 and is based on the assumption that a uniform layer of tissue is ablated with each pulse. In addition to this ablation, the laser was programmed to ablate a profile adjustment to correct for steep central islands. However, for these studies, this additional ablation was removed to study the effect of spatially varying ablation rates with an ablation algorithm which assumes the ablation of tissue to be constant. Removing this additional ablation made software version 4.10 similar to earlier versions which assumed corneal ablation rates to be uniform across the exposed tissue. For all ablations, we assumed the corneal ablation rate to be 0.27 µm per pulse for a fluence of 160 mJ/cmp 2.
Prior to ablating the corneas, the system was calibrated by ablating plastic. Both a 6.0-mm, -4.00 D calibration lens and a 6.0-mm by 50-µm deep flat profile were ablated in plastic calibration cards provided with the Vision Key System. The plastic ablations were read in a Nikon OL-7 lensmeter (Nikon, Inc., Torrance, Calif) and scanned with a Sloan Dektak II profilometer (Veeco Instruments, Inc., Santa Barbara, Calif) prior to ablating the corneas. This profilometer uses the displacement diamondtipped stylus to measure the shape of ablations in flat plastic. Ablations made in plastic after corneal ablations were also measured to ensure that the laser beam did not degrade.
Corneal Preparation and Ablation
Enucleated porcine eye study- Porcine eyes obtained from a local abattoir were kept on ice until use, 4 to 8 hours after slaughter. Prior to epithelial debridement, corneas were visually inspected for surface defects and clarity. The corneas were mechanically debrided of epithelium, exposed to room air for 3 minutes, and the globe placed in a holder. The pressure of the anterior chamber was kept constant by a reservoir 3 to 4 inches, above a needle inserted into the anterior chamber at the corneal limbus. Pre-and post-ablation images of the corneas were obtained with the profilometer. Initially, multiple images of the pre- and postablation surfaces were taken to verify the accuracy and reproducibility of the images, but this redundancy was abandoned because of the consistency and reproducibility of the images.
The laser was programmed to ablate tissue with either PRK or phototherapeutic keratectomy (PTK) ablations, and a new eye was used for each ablation. To compensate for individual variability among the non-living eyes, four eyes were ablated with each instrument setting and the average profile obtained. The ablation pattern expanded from 0.6 mm to a maximum of 6.0 mm during the PRK ablations. We performed four separate enucleated porcine eye substudies for the following purposes:
1) To deteraiine the effect of dioptric correction on the ablated profile, the treatment was set to 6.0 mm and -3.00, -6.00, or -9.00 D.
2) To measure the depth and spatial variance of fixed diameter ablations, the PTK program was set to an ablation diameter of 2, 3, 4, 5 or 6 mm with no transition zone. The ablation depth was programmed to 40 µta and the laser paused for 1 to 2 seconds to capture images at ablation depths of 10, 20, and 40 µm..
3) To measure the evolution of steep central islands during the -6.00 D PRK ablations, the laser was stopped at an aperture opening of 2, 3, 4, 5 and 6 mm for 1 to 2 seconds to capture the corneal image.
4) To test the hypothesis that steep central islands can be corrected by modifying the ablation algorithm, we used the fixed diameter ablation profiles from #2 above to re-engineer the ablation algorithm to produce a spherical ablation on the enucleated porcine eyes. This new ablation algorithm is described in the Table, and consists of five annular treatment zones of varying programmed dioptric correction. In the Table, the annular ablation zone is the region ablated with the programmed spherical correction. The programmed correction is the refractive change which would have occurred in the annular zone if corneal ablation rates had been constant at 0.27 µm per pulse. For the example -6.00 D correction, the region from 0 to 1.5 mm received a programmed correction of -14.60 D, while the region from 4.5 to 6.0 mm received a programmed correction of -4.50 D.
Distribution of Programmed Laser Treatment for a Re-engineered 6.0 mm, -6.00 D Ablation
Live rabbit eye study- Pigmented rabbits at least 12 weeks old and weighing 2 to 3 kg were anesthetized with subcutaneous ketamine 5 mg/kg and xylazine 50 mg/kg. The eyes were proptosed, topical proparacaine 0.5% instilled, and the corneal epithelium removed with a blunt spatula. The corneas were then profiled and ablated with spherical 6.0 mm, -6.00 D ablations in the same manner as in the enucleated porcine eyes. The animals were subsequently euthanized in accordance with ARVO guidelines for animal care.
Profilometer- Figure 2A shows the measured and theoretical elevation difference (µm) between the spherical and 35.00/43.00 D bicurve test surfaces. The measured difference (solid line) between the surfaces is within a few microns of the theoretical difference (dotted line), which is more than adequate to detect a 3-mm by 3.00 D steep central island. This result is representative of the accuracy of the profilometer. Although a single video frame from both the pre- and post-ablation surfaces was used to measure the ablation shape, this shape was shown to be highly reproducible with repeated measurements taken before and after our ablation experiments. The theoretical curve shown is deeper than a -8.00 D PRK surface because the test surfaces, which were measured with a keratometer, use the index of refraction of the aqueous humor, 1.3375, rather than the index of refraction of the cornea, 1.377 (Munnerlyn et al3).
Excimer laser system- The excimer laser produced the desired ablations in plastic. The 6.0 mm, 50-µm ablations were flat to within ± 5 µm by profilometry using a diamond-tipped stylus and flat to within -0.50 D when read on a lensmeter. The 6.0mm, -4.00 D ablations were within ± 0.25 D by lensmeter readings and accurate to ± 3 µm. by profilometry with a diamond-tipped stylus (Fig 2B).
Enucleated porcine study- Compared to anticipated profiles, the ablations shown in Figures 3-5 were over-ablated (more than the expected ablation rate) peripherally and under-ablated centrally. For the 6.0-mm PRK ablations shown in Figure 3A, this effect is present for each dioptric correction. The amount of central under-ablation increases with the PRK correction. Normalizing the measured ablation depth to the maximum assumed ablation depth is an effective way to compare the effect of ablation depth on ablation shape, as is shown in Figure 3B. This figure shows that the depth of ablation has little effect on the ablation shape, and the percent error in terms of ablation shape remains constant.
The spatial variance of stromal ablation rates is demonstrated with the 2 to 6 mm fixed diameter intended 10 µm ablations shown in Figure 4A. The peripheral over-ablation follows the edge with the decreased aperture. Also, the central ablation increases as the aperture closes. The peripheral ablation depths for the fixed diameter 10 µm ablations are 15 µm ± 2 µm. This variability in ablation at the periphery cannot be explained by the curvature of the cornea. For a 43.00 D curved surface with a 7.85 mm radius of curvature, the irradiance and predicted ablation rate are decreased by 8% and 11% respectively at the edge of a 6 mm ablation relative to the center. The variability in peripheral ablation rates is most likely caused by variability in water content among the eyes.
The fixed diameter 3 and 6 mm normalized ablation profiles shown in Figure 4B were obtained by dividing the measured ablation profile by the intended ablation depth of 10, 20 or 40 µm. The normalized profiles show no change in ablation shape with increasing ablation depth. The invariability of fixed diameter ablation shape with depth is consistent with the PRK ablations (Fig 3). At the fixed diameter ablation edges, the measured profile is not as steep as the intended ablation. This artifact is caused by vignetting of the ablated surface by the adjacent unablated surface and is more pronounced for deep 3 mm ablations.
Figure 2: Instrumentation validation. A) Profile obtained by subtracting the image of a 35.00 and 43.00 D bicurve from a 43.00 D sphere. B) Plastic ablation of -4.00 D lens. Theoretical profiles are represented by dashed lines; measured profiles by solid lines.
Given that a PRK ablation for myopia results from the summation of a series of discrete ablations of increasing diameter, the resulting shape of a PRK ablation should depend on the ratio of large diameter pulses to small diameter pulses. Since the small diameter pulses show over-ablation and less underablation, one would expect small diameter PRK ablations to be over-ablated, but more spherical. Alternatively, large diameter PRK ablations should present a preponderance of large diameter ablations with central under-ablation and peripheral overablation. Consistent with the spatial variance of stromal ablation rates, the 2-mm PRK shown in Figure 5 is spherical, but over-ablated by 50%. Also, the additional pulses occurring while the ablation diameter increases from 2 to 6 mm cause the finished PRK ablation to be under-ablated centrally and over-ablated peripherally.
To better assess the sphericity of 2 to 6 mm diameter spherical ablations, the ablation profile can be scaled to both the ablation diameter and maximum depth of ablation. The ablation profiles shown in Figure 5 have been normalized by this method in Figure 6. Consistent with other data, ablation rates near the edge of the ablation for all diameters remain nearly constant and 50% greater than the assumed ablation rate of 0.27 µt? per pulse. The central depth of ablation consistently decreases with increasing ablation diameter. For example, the 5 mm PRK is under-ablated by 13%, whereas the 6 mm PRK is under-ablated by 22%.
Once the ablation profiles for a uniform beam of varying diameter are known, the shape of the ablation can be corrected by modifying the laser ablation algorithm to compensate for the spatial variance of tissue ablation. The validity of this concept is clearly demonstrated with the 6 mm, -6.00 D enucleated porcine eye ablation (Fig 7). The central under-ablation and peripheral over-ablation are substantially reduced, and the ablation is spherical to within + 5 µm.
Figure 3: Enucleated porcine eye with. 6 mm PRK ablations of -3.00, -6.00. and -9.00 D. A) Measured ablations compared to intended. B) Measured (solid lines) ablations normalized to the maximum intended (dashed lines) ablation depth.
Live rabbit eye study- The average ablation profile of four rabbit eyes with 6 mm, -6.00 D ablations is shown in Figure 8. The ablation profile was closer to the theoretically predicted curve for live rabbit eyes than for the enucleated porcine eyes. The peripheral over-ablation and central underablation are less pronounced for the live rabbit eyes. This difference in ablation shape may have been caused by the increased water content of the enucleated porcine eyes.
The hypothesis that corneal stromal ablation shape exactly matches the laser beam profile is shown by this study to be invalid. This study demonstrates that corneal ablation rates are not uniform across a uniform beam and vary with ablation diameter. Ablations can be made spherical by modifying the ablation algorithm to correct for the spatial variance of tissue ablation rates. Ablation algorithms using a uniform laser beam which do not compensate for this spatial variance will produce central underablation and peripheral over-ablation. However, it is the shape of the fully healed, epithelialized cornea, not the initial ablated profile, which is of primary importance in determining the final outcome of PRK
This study provides convincing evidence that the majority of clinically significant steep central islands are caused by the central under-ablation and peripheral over-ablation that we have observed. Our experiments also lead us to believe that patients receiving greater dioptric corrections, larger ablation diameters, and transepithelial ablations which do not correct for the spatial variance of ablation rates are at greater risk of developing steep central islands.
Figure 4: Enucleated porcine eye ablations of 2, 3, 4, 5, and 6 mm. A) Intended 10 pm ablation profiles. B) Intended 10, 20, and 40 pm ablations (dashed lines) normalized by intended ablation depth (3 and 6 mm).
The extent to which the spatial variance of ablation observed with our laboratory study represents the PRK ablation with patients is not clear. Following PRK, steep central islands have been documented by corneal topography, and it is extremely likely that spatial variance occurs which is similar to that found in our study. The live rabbit eye profiles show less spatial variance in ablation than the enucleated porcine eye profiles. This difference in spatial variance of ablation may have been caused by the increased hydration of the porcine corneas, which had typically swollen 100 µt? prior to ablation. This difference in spatial variance of ablation indicates that the ablation algorithm used to produce spherical enucleated porcine eye ablations may not be appropriate for PRK patients.
The underlying cause of the variable ablation rates across a uniform beam is not clear, although many hypotheses have been made to explain the steep central island phenomenon. The ablation process is complex, and several factors are likely to contribute to the phenomenon. Explanations of the spatial variation in ablation rates include: 1) differential hydration1921,23; 2) redeposition of ablated material as it lifts from the surface12· 29; 3) the spallation model for ablation30; 4) dependence upon the proximity to an unablated area; and 5) a persistent ablation fog overlying the stromal surface.
Figure 5: Steep central island evolution during -6.00 D PRK. Enuclated porcine eye ablation profiles taken with increasing diameter from 2 to 6 mm in 1 mm increments. Intended ablations represented by dashed lines and measured by solid lines.
Figure 6: Effect of PRK ablation diameter on ablation shape. Ablations from Figure 5 normalized to the maximum intended (dashed lines) ablation depth and diameter
Several investigators1921,23 have suggested the possibility that steep central islands are created by regional differences in corneal hydration during laser ablation. This hypothesis is based on the clinical observation that the cornea appears dry at the edge of the ablation, and moist centrally. The depth of penetration of the laser pulses is known to depend on tissue hydration31, and moist cornea ablates more slowly than dry cornea.8 For a fluence of 160 mJ/cm2, a typical ablation rate is 0.27 µm per pulse. With corneas desiccated by dry nitrogen, the ablation rates approach 0.5 µm. per pulse (unpublished live rabbit eye studies at Louisiana State University and unpublished enucleated bovine eye studies at VISX). However, when hydrated nitrogen (92% relative humidity) is blown over the cornea, the ablation rates return to 0.27 µta. per pulse (unpublished enucleated bovine eye studies at VISX). Clearly, hydration of the corneal stroma is a critical factor influencing ablation. Unfortunately, this model cannot explain why we have observed a similar, albeit lesser, effect in PMMA plastic, which is not hydrated (Unpublished plastic ablation studies, VISX, Inc.)
The redeposition of ablated material is likely to occur during the ablation of both PMMA and cornea. Since the cornea is approximately 80% water, it would likely appear dry at the edge and moist centrally if ablated cornea were to be redeposited centrally. This appearance of central moisture is known to occur during clinical ablation. As the corneal stroma and PMMA are ablated, very small particles are ejected from the surface at high velocities.29,32 Some of this material probably accumulates on the treated area and absorbs laser energy from subsequent pulses,12 thus reducing the amount of tissue ablated with each subsequent pulse. This redeposition model is also supported by the computer simulations of Sibold and Urbassek33, which predict greater amounts of ablated material redeposition with increasing ablation diameters.
Figure 7: Effect of central island software on 6 mm, -6.00 D enucleated porcine eye ablation profiles. (Intended ablation - fine dashed line, modified ablation algorithm - coarse dashed line, original ablation algorithm- solid line).
Figure 8: Live rabbit eye 6 mm, -6.00 D ablation profiles (intended ablation - dashed line, measured ablation- solid line).
The calibration plastic used with the Vision Key system was developed to avoid problems associated with debris during the ablation of PMMA. Ablations with this new material have been shown to better match the laser beam irradiance pattern than ablations with PMMA (unpublished studies at "VISX), and little, if any debris is present after the ablation. Because the laser beam was checked both before and after the corneal ablations with this new material, we have ruled out the possibility that the measured spatial variance in tissue ablation was been caused by the laser beam.
Spallation is a known mechanism for the ablation of tissue30 and plastic. This ablation mechanism is characterized by a Shockwave which causes the breaking and ejection of material. This model for ablation may produce more pronounced ablation near an interface such as an ablation edge, which would partially explain our experimental findings. However, this model does not address other aspects of the ablation process, such as differential hydration, ablated material redeposition, and occlusion of the laser beam by residual fog from the previous laser pulse.
By analyzing video frames, we have determined that most of the plume is ejected far above the ablated area and removed as it passes by the aspirator, 1 3/4 inches above the ablation. This removal process occurs prior to the next laser pulse. However, a small residual plume, or fog, remains adjacent to the ablation until the next pulse. This fog is barely visible in the operating microscope due to the back scattered viewing of the ablated area and plume. However, this residual plume appears as a dark cloud with our profilometer because it is being viewed with transmitted light passing obliquely over the cornea. The fog does not appear until the ablation diameter is 2 to 3 mm or more, which is consistent with the diameter at which the central under-ablation becomes apparent. However, in a separate set of experiments, when we blew the fog away with moist nitrogen, the center was still under-ablated, but to a lesser extent. Therefore, attenuation of the laser beam by the residual plume from the previous laser pulse can only partially explain our results.
Concerns regarding debris prompted VISX to initially blow nitrogen over the cornea. This purging provides a uniform corneal surface which appears gray under the operating microscope. However, studies have shown that drying the cornea during the ablation process produces rougher ablations in tissue.5·6 These results and the possibility of increased postoperative haze led to controlled clinical evaluations of the nitrogen purge at selected sites.
Although it has been observed clinically that the incidence of steep central islands increased when the nitrogen purge was abandoned, the surgical method of choice has been to not use nitrogen purge and keep the cornea moist. Lin and associates19 did not observe steep central islands in eyes in which nitrogen debris removal was used. In the absence of effluent removal using nitrogen gas, Krueger and associates22 found over 64% of patients had steep central islands that were present at 1 month or beyond, compared to only 20% when gas blowing was employed. However, removal of the nitrogen purge improved the outcome of the surgery, resulting in faster reepithelialization, less haze, and greater precision in the laser correction achieved.10,12
Despite the presence of clinically significant steep central islands immediately following PRK, it has been observed that, with time, the steep central islands tend to dissipate, and any concomitant visual aberrations tend to improve.19,22 Improvement of visual symptoms associated with steep central islands may be explained in many cases by corneal smoothing occurring with epithelial cover and stromal repair. Thus, although some steep central islands can create clinically significant visual anomalies in the early postoperative period, most are eliminated spontaneously.
Modifying the ablation algorithm to accommodate the spatial variance of corneal ablation has produced spherical enucleated eye PRK ablations. By analyzing the postoperative topography of PRK patients, we have modified the ablation algorithm to produce spherical ablations in situ. This software should reduce or eliminate the incidence of visually significant corneal topographic abnormalities, accelerate the recovery of spectacle-corrected vision, and further enhance the safety and efficacy of excimer PRK procedures.34,35
1. Trokel S, Srinivasan R, Braren B. Excimer laser surgery of the cornea. Am J Ophthalmol 1983; 96:710-715.
2. Marshall J, Trokel S, Rothery S, Krueger R. Photoablative reprofiling of the cornea using an excimer laser: Photorefractive keratectomy. Lasers Ophthalmol 1986;1:2148.
3. Munnerlyn C, Koons S, Marshall J. Photorefractive keratectomy: a technique for laser refractive surgery. J Cataract Refract Surg 1988;14:46-52.
4. Fantes F1 Waring G. Effect of excimer laser radiant exposure on uniformity of ablated corneal surface. Lasers in Surg and Med 1989;9:533-542.
5. Campos M, Cuevas K, Garbus J, Lee M, McDonnell P. Corneal wound healing after excimer laser ablation. Effects of nitrogen gas blowing. Ophthalmology 1992;99:893-897.
6. Krueger R, Campos M, Wang X, Lee M, McDonnell P. Corneal surface morphology following excimer laser ablation with humidified gases. Arch Ophthalmol 1993;111:1131-1137.
7. Krueger R, Trokel S. Quantitation of corneal ablation by ultraviolet laser light. Arch Ophthalmol 1985;103:17411742.
8. Dougherty P, Wellish K, Maloney R. Excimer laser ablation rate and corneal hydration. Am J Opthalmol 1994;1 18:169176.
9. Campos M, Wang X, Herzog L, Lee M, Clapham T, Trokel S, McDonnell P. Ablation rates and surface ultrastructure of 193 zun excimer laser keratectomies. Invest Ophthalmol Vis Sci 1993;34:2493-2500.
10. Salz J, Maguen E, Nesburn A, Warren C, Macy J, Hofbauer J, Papaioanna T, Berlin M. A two-year experience with excimer laser photorefractive keratectomy for myopia. Ophthalmology 1993;100:873-882.
11. Tengroth B, Epstein D, Fagerholm P, Hamberg-Nystrom H, Fitzsimmons T. Excimer laser photorefractive keratectomy for myopia. Clinical results in sighted eyes. Ophthalmology 1993;100:739-745.
12. Piebenga L, Matta C, Deitz M, Tauber J, Irvine J, Sabates F. Excimer photorefractive keratectomy for myopia. Ophthalmology 1993;100:1335-1345.
13. Les Jardins S, Aucun F, Roman S, Burtschy B. Results of photorefractive keratectomy on 63 myopic eyes with 6 months minimum follow-up. J Cataract Refract Surg 1994;20:223-228.
14. Kim J, Hahn T, Lee Y, Sah W. Excimer laser photorefractive keratectomy for myopia: two-year follow-up. J Cataract Refract Surg 1994;20:229-233.
15. Talley A, Sher N, Kim M, Doughman D, Corpel E, Ostrov C, Lane S, Parker P, Lindstrom R. Use of the 193 nm excimer laser for photorefractive keratectomy in low to moderate myopia. J Cataract Refract Surg 1994;20:239-242.
16. Heitzmann J, Binder P, Kassar B, Nordan L. The correction of high myopia using the excimer laser. Arch Ophthalmol 1993;111:1627-1634.
17. Klyce S, Smolek M. Corneal topography of excimer laser photorefractive keratectomy. J Cataract Refract Surg 1993;19:122-130.
18. Wilson S, Klyce S, McDonald M1 Liu J, Kaufman H. Changes in corneal topography after excimer laser photorefractive keratectomy for myopia. Ophthalmology 1991;98:1338-1347.
19. Lin D, Sutton H, Berman M. Corneal topography following excimer photorefractive keratectomy for myopia. J Cataract Refract Surg 1993;19:149-154.
20. Parker P, Klyce S, Ryan B, Lindstrom R, Sher N, McDonald M, Lane S1 Carpel E, Ostrov C, Doughman D. Central topographic islands following photorefractive keratectomy. Invest Ophthalmol Vis Sci 1993;34(suppl):803.
21. Levin S, Carson C, Garrett S, Taylor H. The incidence of central islands following excimer laseT refractive surgery. Invest Ophthalmol Vis Sci 1994;35(suppl):2018.
22. Krueger R, Saedy N, McDonnell P. Clinical analysis of steep central islands following excimer laser photorefractive keratectomy (PRK). Arch Ophthalmol 1996;114:377-381.
23. Cochener B, Gallinaro C, Colin J. Central steep islands immediately following excimer photorefractive keratectomy for myopia. Invest Ophthalmol Vis Sci 1994;35(suppl):1740.
24. Kawesch G, Maloney R, Derse M, Waring G, Seiler T. Contour of the ablation zone after photorefractive keratectomy. Invest Ophthalmol Vis Sci 1992;33(suppl):1105.
25. Hersh P, Schwartz-Goldstein B. Corneal topography of phase ITJ excimer laser photorefractive keratectomy: characterization and clinical effects. Ophthalmology 1995;102:963-978.
26. Seiler T, Holschbach A, Derse M, Benedict J, Genth U. Complications of myopic photorefractive keratectomy with the excimer laser. Ophthalmology 1994;101:153-160.
27. Summit Technology, ExciMed UV200LA Laser System User's Manual. Summit Technology Ine, Waltham MA. 1991:5-25.
28. Shimmick J, Munnerlyn C. Corneal analysis with a rectilinear photokeratoscope. Ophthalmic and Visual Optics Technical Digest Series 1992;3:2-3.
29. Srinivasan R, Braren B, Casey K, Yeh M. Ultrafast imaging of ultraviolet laser ablation and etching of polymethylmethacrylate. Appi Phys Lett 1989;55:2790-2791.
30. Dingus R, Scammon R. Gruneisen stress induced ablation of biological tissue. Proc SPIE 1991;1427:45-54.
31. Ediger M, Petit G, Weiblinger R, Chen C. Transmission of corneal collagen during ArF excimer laser ablation. Lasers Surg Med 1993;13:204-210.
32. Puliafito C, Stern D, Krueger R, Mandel E. High-speed photography of excimer laser ablation of the cornea. Arch Ophthalmol 1987;105:1255-1259.
33. Sibold D, Urbassek H. Effect of gas phase collisions in pulsed laser desorption: a three dimensional Monte Carlo simulation study. J Appi Phys 1993;73:8544-8551.
34. Kraff C. Can corneal central islands be ehminated? ISRK Current Research: Refractive and Corneal Surgery Symposium 1994;39.
35. Machat JJ. Preliminary results of the VISX central island factor software. Invest Ophthalmol Vis Sci 1994;35suppl):1740.
Distribution of Programmed Laser Treatment for a Re-engineered 6.0 mm, -6.00 D Ablation