Although laser photoablation will be a key technique for future developments in minimally invasive procedures in general surgery, and in ophthalmic surgery in particular, its physical parameters have not yet been fully investigated.
Photoablation of hydrated tissue is possible with laser light in the far UV (193 to 308 nm) using excimer lasers1"3 or frequency multiplied solid state lasers,4 and in the mid-infrared around 3 µm (HF laser, EnWi laser, EnYSGG).5-8
An essential prerequisite for photoablation seems to be high absorption in the target material. However, photoablation - at least in the far UV - is a nonlinear process; although low energy spectroscopy provides approximate data for photoablation, this cannot be transferred quantitatively to the high energy domain because the photon-atom interactions in the infrared (IR) are not yet fully understood. While with current laser technology, shifts in wavelength are either impossible or limited to a narrow range (eg, Ti: Sapphire laser), the free electron laser can be tuned over a wide range, also allowing variable combinations with pulse duration, energy, and repetition rate.
The goal of this work was to study the basic principles of photoablation dynamically over a wide range of wavelengths, using the advanced free electron laser of the Free Electron Laser Center, Vanderbilt University, Nashville, Tenn,9 as a research tool. In this series of experiments, wavelength was varied between 2.7 and 6.7 µm, while pulse duration was kept constant at 4 µs. The efficiency of the ablation process was studied for different hydrations of standardized gelatin probes as a proven model for the cornea.
Free electron lasers are based upon a high energy electron beam injected into the laser cavity. Electrons, undulated by a series of electromagnetic fields (wiggler), generate electromagnetic radiation emitted in the form of IR light. The wavelengths depend upon the energy of the injected electrons and the amplitude of the undulation. The electrons are recycled and the resulting electromagnetic radiation is amplified in the conventional manner, forming an IR laser beam.
Figure 1: (A) Set-up for perforation experiments with the free electron laser (FEL). For the depth measurements, the power meter was removed. (B) Ablation profile in gelatin (single profile). Depth measurements with confocal laser topometry (fluence 1.27 J/ cmp 2; wavelength 6.2 mm; 10 pulses at 2 Hz repetition rate).
Ablation properties as a function of wavelength were related to corresponding measured absorbencies of the target materials, gelatin, and human corneal tissue.
The data of this study will help to develop a model which could result in a better understanding of the laser-tissue interaction and possibly achieve predictability.
MATERIALS AND METHODS
The study was carried out in September 1992 with the free electron laser at the Free Electron Laser Center at Vanderbilt University, Nashville, Tenn. At that time, wavelengths were tunable between 2.7 and 6.7 µp? with a constant fluence of 1.27 J/cmp 2 at every wavelength. A beam diameter of 1 mm with a reproducible and stable beam profile, rectangular in first approximation, a stable pulse-to-pulse energy, described elsewhere,9 and a laser repetition rate of 2 to 20 Hz were used. Pulse duration was kept constant at 4 µß. Each 4 µß macropulse consisted of a track of 2 ps micropulses with a repetition rate of 2.9 GHz.
All ablation experiments were performed on gelatin discs, a proven model for photoablation in corneal tissue,10 allowing the study of the relevance of variable water content of the target material. The discs were prepared by cooking red gelatin (Oetker, FRG). After drying for 2 days under UV Clear Bank radiation at ambient temperature, discs with a diameter of 10 mm were punched out with a trephine. The thickness of each disc was 140 ± 5 µm, as determined with a mechanical caliper and confocal laser topometer (UBM, Ettlingen, FRG).
Even at minimal hydration (wet/dry weight 1) gelatin still contains about 14% water.
Ablation Rate as a Function of Wavelength
To determine the ablation rate in dry gelatin as a function of wavelength (2.7 to 6.7 µm), two types of experiments were performed:
Perforation Experiments. The ablation rate was calculated by counting the number of laser pulses needed to perforate the gelatin disc. Perforation was measured using a power meter (Gentec ED 500, Canada) as shown in Figure 1. A PC was used to count the number of applied laser pulses, allowing at the same time the triggering of the free electron laser by means of specially developed software.
Measurements of Ablation Depth. Defined numbers of laser pulses (10, 20, 40, 80 pulses) were applied sequentially to the gelatin probe. The same set-up as before was used, with the exception of the power meter (Fig IA). After laser irradiation, the depth of excision on the gelatin discs was measured using confocal laser topometry (Fig IB). For each series of pulses, absolute ablation depth was measured. The quotient of the ablation depth and the number of laser pulses leads to the ablation depth per pulse at a given wavelength; its mean value for each wavelength leads to a mean ablation rate per pulse at a given wavelength. These data underwent regression analysis.
Figure 2: (A) Absorption spectrum of human cornea and gelatin measured by surface spectroscopy. (B) Results of the ablation experiments measured by perforation and depth measurements. Fluence was 1.27 J/cmp 2 at 2 Hz. * marks ablations with 3.5 J/cmp 2 (only perforation experiments).
Figure 3: Ablation rate at 6.2 µm as a function of fiuence for perforation and depth measurements. Repetition rate was 2 Hz at a beam diameter of 1 mm.
Ablation Rate as a Function of Fluence and Repetition Rate
Using the above-mentioned techniques, the ablation rate was studied as a function of fluence (0.8 to 1.75 J/cmp 2) as the laser repetition rate was varied between 1 and 20 Hz at a wavelength of 6.2 µm. Again, perforation counts and depth measurements were taken.
Ablation Rate as a Function of Hydration
Since water is the main absorbent in the mid-IR, ablation rate as a function of hydration of the gelatin (wet/dry weight ratio) was studied; measurements were taken at a wavelength of 6.2 µm and a constant fluence of 1.27 J/cmp 2. For controlled hydration, the gelatin discs were kept in saline solution for different time intervals. After storage in a wet chamber for 20 minutes to achieve a steady state, the discs were weighed to assess hydration. In this manner, hydrations between 1 and 4.8 wet weight/dry weight were obtained. By checking weights (accuracy 0.01 mg), it was confirmed that disc hydration remained constant for more than 25 minutes, while measuring time was less than 2 minutes.
Ablation rates were assessed by perforation experiments only, since there is no way of storing the hydrated gelatin for longer periods. Confocal laser topometry is not feasible here.
Absorption Spectrum of Gelatin and Human Cornea
To compare the results of the photoablation experiments as a function of wavelength, absorption spectroscopy of gelatin and the human cornea was performed using a surface spectrometer (Perkin Elmer 1710) in the same wavelength range (2.7 to 6.7 µm).
Figure 2A describes the ablation rate as a function of wavelength and relates it to the absorption spectra of the human cornea (solid line) and gelatin (dotted line), measured with a surface spectrometer; Figure 2B describes ablation rates in gelatin at a constant fluence of 1.27 J/cm2. The two curves for the ablation rates represent the ablation rate of the perforation experiments (solid line) and the measurements with the confocal laser topometer (dotted line).
Figure 4: Ablation rate at 6.2 µm as a function of laser repetition rate. Fluence. was 1 .27 J/cmp 2.
Figure 5: Ablation rate at 6.2 (µm as a function of hydration in gelatin. Fluence was 1.27 J/cmp 2; pulse repetition rate was 2 Hz.
The ablation rate is parallel with the absorption values of the target material. It is maximal around 3 and 6.2 µm. At minimal absorption, only a nonablative thermal effect is observed; no photoablation is observed. At 6.4 µm - an absorption peak of protein in corneal tissue - no significant influence on the ablation rate was found. (Gelatin does not show this absorption peak.)
In Figure 3, ablation rates are shown as a function of fluence (logarithmic scale) for the absorption maximum at 6.2 µmp 2. Laser repetition rate was 2 Hz. Extrapolation indicates an ablation threshold of 0.7 ± 0.05 J/cmp 2 for the perforation experiments and 0.55 ± 0.08 J/cmp 2 for the depth measurements.
Ablation rate (6.2 µm wavelength, 2Hz repetition rate, fluence 1.27 J/cmp 2) as a function of the laser repetition rate shows a continuous increase, possibly ending in a plateau for higher repetition rates at >20 Hz pulse repetition rate (Fig 4). This needs to be confirmed when pulse repetition rates >20 Hz become technically available.
Ablation rate as a function of hydration (Fig 5) shows an increase in the ablation rate up to 2.3 (wet/dry weight). Higher hydrations then decrease the ablation rates: These data allow the translation of the ablation rate to any state of target material hydration or to normally-hydrated corneas.
With the free electron laser, it is possible for the first time to perform ablation experiments in the IR by tuning wavelengths over a wide range from 2.7 to 6.7 µm, in combination with variable fluences and pulse lengths, at a sufficient pulse-to-pulse stability and a reproducible beam profile. The accuracy of a selected wavelength is better than 3 nm (Free Electron Laser Center data). However, the output energy of this free electron laser system depends upon its wavelength (eg, the output energy decreases with shorter and longer wavelengths). To compare the results of the ablation experiments at several wavelengths, a constant fluence of 1.27 J/cmp 2 and a beam diameter of 1 mm were chosen. This fluence was achieved for all selected wavelengths.
Gelatin is a feasible and proven substitute for corneal tissue.10 Both are water-containing polymers with an. almost identical absorption spectrum (Fig 2A), except for a protein absorption peak which the cornea typically shows at 6.45 µm. Gelatin has been used successfully in the development of UV photoablation10 for calibration and measurement purposes. Its water content of about 14% (wet/dry weight = 1) is high enough to show the typical water absorption bands (Fig 2A); it is thus justified as a corneal substitute, because even in its driest state, water is still the prevailing absorbent.
Figure 2B shows that ablation rates as a function of wavelength correlate well with the absorption spectrum of the target material: high absorption corresponds to a high ablation rate. First orientation experiments (Fig 2B) showed that ablation is also possible at wavelengths with low absorption when fluence is increased (3.5 J/cmp 2). This will be studied in detail in future experiments.
The offset between the two ablation rate curves (Fig 3) indicates that perforation experiments are less reliable, especially for the last few pulses - when first breaks in the gelatin are seen - the fluttering of the disc bottom may influence the effective ablation rate, Videographic analysis revealed that the bottom may crack when still 10 µm thick. Cracks are presumably caused by shockwaves, observed during photoablation.11 Therefore, results of confocal laser topometry measurements are more reliable because they quantify the true ablation depth.12
The ablation rate as a function of frequency increases at a constant fluence (1.27 J/cmp 2). The curve obtained by regression analysis is the one which best fits the measuring points and error bars. This is in agreement with earlier findings of UV excimer photoablation,13 where a 3% increase of the ablation rate was found over the same range of pulse repetition rates (1 to 20 Hz). For the 6.2-micrometer wavelength, the 60-percent increase described here indicates that the pulse repetition rate is an important determining element of photoablation in the IR.
Furthermore, earlier studies showed that increased repetition rates lead to a temperature increase in adjacent tissue for the Er:YAG's 2.94micrometer wavelength14 as well as for the excimer's 193-nanometer radiation.15 A possible explanation is that higher temperatures caused by higher pulse repetition rates together with the long thermal cooling time of corneal tissue (several seconds),14 may lead to a change within the target material beyond mere dehydration.
Target hydration itself, however, also causes different ablation rates, as demonstrated by Figure 5.
For the 6.2-micrometer wavelength, Figure 5 shows an increase of the ablation rate as a function of hydration up to a 2.3 wet weight/dry weight with a drop at higher hydrations. The same has been described for the 2.94 µm ErfYAG laser with a maximum at a hydration of 2.3 in gelatin.16
While all experiments were performed at a hydration of 1 (wet/dry weight) in gelatin, Figure 5 allows the transference of data for any target material hydration, thus making ablation rates predictable for the cornea's typical physiological hydration of 3.3.
Figure 3 shows a squared increase of the ablation rate as a function of fluence, where a linear increase could be expected for the IR (linear absorption process). This conclusion, however, is not definite since the 1.5 and 1.7 J/cmp 2 measurements show large error bars and higher fluenees cannot yet be provided by the free electron laser. Should this squared increase also be confirmed for higher fluenees, this would be a first indication that nonlinear processes are present during 6.2-micrometer photoablation. From Figure 3, the ablation threshold can be defined at 0.68 ± 0.05 J/cmp 2 for the perforation experiments and at 0.52 ± 0.08 J/cmp 2 for the depth measurements.
It has been speculated that the free electron laser's pulse characteristics, with macropulses (4 µß in our set-up) composed of 2 ps micropulses at GHz repetition rate, generate a different type of photoablation which cannot be compared to typical solid state laser pulses. These ablation rate experiments reveal no substantial difference between the ablation process with the free electron laser and the Er:YAG laser.7,8,13,14 The thermal adverse effects are compared in another paper.17
The question as to the relevance of pulse duration for photoablation in general is of major importance and has to be studied as soon as pulse length variation becomes technically available.
The goal of this experiment was not to use the free electron laser as a tool for clinical application, but to determine wavelengths at which sufficient photoablation occurs. As well as at the known 3 -micrometer wavelength, this is the case at around 6.2 µm, For the first time, photoablation was performed successfully at this wavelength, possibly a new option for future clinical application. Further study is now required to provide appropriate solid state laser sources for practical clinical use.
1. Trokel SL, Srinivasan R, Braren B. Excimer laser surgery of the cornea. Am J Ophthalmol. 1983;96:710.
2. Berlin M, Bende T, Schultz U, Martinez M, Seiler T. Photoablation with 308 nm in human sclera. Invest Ophthalmol. 1989;30:281.
3. Seiler T, Bende T, Wollensak J. Einsatz von fernem UV-Licht zur Photoablation der Hornhaut. Fortschr Ophthalmol. 1986;83:556.
4. Feld JR, Lin CP, Woods WJ, Zenzie HH, Rines GA, Moulton PF, Puliafito CA. Cornea ablation studies at wavelength between 205 and 225 nm using a tunable solid state laser. Invest Ophthalmol Vis Sci. 1992;38(suppl):1105.
5. Loertscher H, Mandelbaum S, Parrish RK, Parel JM. Preliminary report on corneal incisions created by a hydrogen fluoride laser. Am J Ophthalmol. 1986,102:217.
6. Feuerstein M, Seiler T. Corneal ablation by mid-infrared laser. In: Marshall J, ed. Laser Technology in Ophthalmology. Amsterdam: Kugler Publications; 1988.
7. Bende T, Kriegerowski M, Seiler T. Photoablation in different ocular tissues performed with an erbium:YAG laser. Lasers Light Ophthalmol. 1989;2:263.
8. Bende T, Seiler T, Wollensak J. Photoablation mit dem Er:YAG-laser an okulären Geweben. Fortschr Ophthalmol. 1991;88:12.
9. Brau C. The Vanderbilt University Free-Electron Laser Center. Nuclear Instruments & Methods in Physics Research , A 318. Elsevier Science Publisher B.V.; 1992:38-41.
10. Bende T, Matallana M, Seiler T. Kalibrierung des 193 nm Excimerlaser Strahls. Biomedizinische Technik. 1990;35(suppl):14.
11. Kermani O, Lubatschowski H. Struktur und Dynamik photoakustischer Schockwellen bei der 193 nm Excimer Laser Photoablation Fortschr. Ophthalmology. 1991;88:748-753.
12. Bachmann W, Jean B, Bende T, Seiler T, Hibst R, Thiel HJ. Silicone cast method for quantification of photoablation. J Refract Corneal Surg. 1992;8:363.
13. Kriegerowski M, Bende T, Bachmann W, Oltrip T, Jean B. Corneal silicone cast in vivo and in vitro for photorefractive keratectomy. Invest Ophthalmol Vis Sci. 1993 ;4(suppl): 1251.
14. Bende T, Jean B, Matallana M, Seiler T. Thermal grathents in the cornea during photoablation with the Er:YAG laser. Laser Light Ophthalmol. 1992;5:79.
15. Bende T, Seiler T, Wollensak J. Side effects in excimer laser surgery: corneal thermal grathents. Graefes Arch Clin Exp Ophthalmol. 1988;226:277.
16. Freund S. Photoablation with an EnYAG laser (2.94 µm) as a function of target hydration. Dissertation thesis, Eye Clinic, Free University, Berlin 1993.
17. Bende T, Jean B, Seiler T, Brau C. Photoablation and thermal damage in porcine cornea using an FEL between 2.7 and 6.4 µm. J Refract Corneal Surg. In press.