bptwlfcdvi

hi

Hollow-glass waveguide delivery of an infraredfree-electron laser for microsurgical applications

Jin H. Shen, James A. Harrington, Glenn S. Edwards, and Karen M. Joos

The purpose of this research is to deliver free-electron-laser ~FEL! pulses for intraocular microsurgery.The FEL at Vanderbilt University is tunable from 1.8 to 10.8 mm. To deliver the FEL beam we used ametallic-coated hollow-glass waveguide of 530-mm inner diameter. A 20-gauge cannula with a minia-ture CaF2 window shielded the waveguide from water. Open-sky retinotomy was performed on cadavereyes. The system delivered as much as 6 3 105 W of FEL peak power to the intraocular tissues withoutdamage to the waveguide or to the surgical probe. © 2001 Optical Society of America

OCIS codes: 140.2600, 140.3600, 170.1020, 170.3890, 170.4470, 230.7370.

bgw

mm

1. Introduction

The Mark-III free-electron laser ~FEL! at VanderbiltUniversity operates in the 1.8–10.8-mm region. Itproduces macropulses of ;5-ms duration with a rep-etition rate of as much as 30 Hz. Each macropulsecomprises a train of 1-ps micropulses repeating at 3GHz.1 By tuning the infrared laser wavelength tothe absorption bands of the tissue, i.e., the molecularvibrational modes ~for example, 2.94– and 6.1-mmwater bands, 6.0-mm Amide I band, 6.45-mm Amide IIand, and 7.7-mm Amide III band!, one can identifyotential optimal wavelengths for specific laser–issue interactions. Ocular tissue ablation studiesith the FEL have identified several useful wave-

engths at which surgical procedures can be per-ormed.2,3 For example, the 6.45-mm wavelength isapable of ablating a retina with minimal collateralamage. For FEL applications in surgery, the de-elopment of a useful laser delivery system is a keyssue.

Various infrared fibers and hollow waveguidesave been studied that can deliver high-peak-power

nfrared lasers. Examples include fluoride glass fi-

J. H. Shen ~jin-hui.shen@mcmail.vanderbilt.edu! and K. M. Joosare with the Department of Ophthalmology and Visual Sciences,Vanderbilt University, Nashville, Tennessee 37232. J. A. Har-rington is with the Fiber Optics Materials Research Program,Rutgers University, Piscataway, New Jersey 08855. G. S. Ed-wards is with the Department of Physics, Duke University,Durham, North Carolina 27708.

Received 1 June 2000; revised manuscript received 10 October2000.

0003-6935y01y040583-05$15.00y0© 2001 Optical Society of America

ers, single-crystal sapphire fibers, chalcogenidelass fibers, polycrystalline fibers, and hollowaveguides.4 Although they offer some ability to de-

liver different infrared wavelengths, each of them hasone or more limitations. Problems include brittle-ness, limited mechanical strength, water solubility,toxicity, sensitivity to UV exposure, low temperature-damage threshold, and low laser-damage threshold.It was not until recently that a flexible, small-diameter waveguide that would withstand high laserpeak power was developed that provided sufficientperformance for medical use.5 We describe a micro-surgical device based on a hollow-glass waveguidethat can deliver FEL energy for potential internalmicrosurgical applications.

2. Materials and Methods

A. Characterization of Hollow-Glass Waveguides

Hollow-glass waveguides ~HGW’s! that have internaletal and dielectric films are currently one of theost attractive means of delivering Er:YAG and CO2

laser power.5,6 A HGW, usually with a metallic filmof Ag and a single dielectric film of AgI or AgBr, havelow loss5 and an output beam profile of high spatialpurity7 as a result of the smooth inner surface and thehigh uniformity in the cross section of the glass tub-ing used as the base material. The structure of atypical HGW is shown in Fig. 1. The theory describ-ing losses in circular hollow waveguides was thor-oughly explored by Marcatili and Schmeltzer in1964.8 Their study was followed by that of Miyagiand Karasawa, who calculated bending losses inthese structures.9 The results of the latter studygive two important relationships for the attenuation

1 February 2001 y Vol. 40, No. 4 y APPLIED OPTICS 583

t

tfw

AcAooi~dehn2wdau

aedpwcTc

wUcot

Trrsw1pTf

tsfo

5

coefficient, a, that govern the losses in hollow guides;hat is,

a }l2

a3 , a }1R

, (1)

where a is the bore diameter and R is the bendingradius. These relationships show that thewaveguide losses will increase dramatically as thebore size is reduced and that there is an additionalloss with bending that increases as the inverse of thebend radius. In addition, hollow waveguides havethe lowest loss for the lowest-order waveguide modes.Higher-order modes have higher loss, and therefore,even though the core diameter is often more than 100times greater than the wavelength, hollowwaveguides tend to propagate near single-mode en-ergy because the higher-order modes are rapidly at-tenuated.

HGW’s are fabricated by a liquid-phase depositiontechnique.10 In the deposition process, first an Agfilm is deposited by a conventional electrolysis plat-ing method similar to that which is used to make Agmirrors. Then AgI film is formed by flowing iodinesolution inside the tubing. The AgI is a protectivecoating for the Ag layer. Also, the AgI thin dielectricfilm–generated interference effect creates the peaksin the transmission spectrum. By adjusting thethickness of the AgI film, one could optimize thewaveguide for a specific wavelength. Figure 2shows the measured spectral loss for 700-mm-boreHGW’s of 50-cm length.

Fig. 1. Structure of a typical HGW, showing Ag and AgI filmsdeposited inside silica tubing.

Fig. 2. Spectral loss for HGW’s designed for long-wavelength op-eration.

84 APPLIED OPTICS y Vol. 40, No. 4 y 1 February 2001

B. Experimental Design

In this experiment we used a hollow waveguide withan inner diameter of 530 mm, and the optimizedransmission wavelength was 6 mm. The maximumull acceptance angle was 3°. The length of theaveguide was 2 m.5 The FEL beam had a single-

mode Gaussian distribution, with a spot size of10-mm diameter. A 150-mm CaF2 lens was used tofocus the FEL beam into a 550-mm diameter spot.

n adjustable diaphragm was placed before the fo-using lens to attenuate and adjust the FEL energy.500-mm-diameter metal pinhole was placed in front

f the entrance of the waveguide to protect the edgesf the waveguide from being damaged by high-ntensity laser-induced plasma during the alignmentFig. 3!. The waveguide was protected by a 3-mm-iameter Teflon tubing, with a SMA connector at onend and a handpiece at the other end. The tip of theandpiece was a disposable 20-gauge thin-wall can-ula. A miniature CaF2 window–lens ~focal length,mm! was fixed in front of the cannula to shield theaveguide from water. Nitrogen gas was intro-uced into the Teflon tube, flowed to the probe tip,nd circulated through the hollow waveguide. Thisnit was sterilized by ethylene oxide gas.The FEL was tuned to a 6.45-mm wavelength and10-Hz repetition rate. We adjusted the incident

nergy from the FEL to the HGW by changing theiameter of the diaphragm. By measuring the out-ut energies from 10-cm- and a 200-cm-long pieces ofaveguide, we determined both the coupling effi-

iency and the transmission loss of the waveguide.he transmission also was measured with a 360°urved waveguide at a radius of 25 cm.

The spot size of the output beam from the probe tipas measured by the scanning knife-edge method.sing different focal lengths of the mini lens would

reate different spot sizes at the focal point. Theptimal spot size was determined by tissue responseo the laser interaction.

Fresh human cadaver eyes were obtained from theennessee Lions Eye Bank. In each case, the ante-ior segment of the eye was removed. The posterioretina, choroid, and sclera were intact for the open-ky experiment. After the vitreous was replacedith a balanced salt solution, the probe was placed–2 mm from the retinal surface. Several visiblearallel FEL incisions were cut, as shown in Fig. 4.he tissue was then fixed in formalin and processed

or histology.

3. Results

The measured coupling loss was 1 6 0.3 dB ~20 65%!. The transmission loss was 2.1 6 0.15 dBym~16.5%!. A 360° bend with a radius of curvature of25 cm reduces transmission by an additional 0.7 60.25 dB ~15 6 5%!. The transmission was beenested continuously for 30 min without change. Theame waveguide was used for at least 1 year at arequency of more than once every other week with-ut significant change of the transmission. During

it

this period the waveguide was gas sterilized severaltimes.

After several trials with cadaver eyes, the surgeondetermined that the optimal focal distance was 2 mm.By the knife-edge technique, the spot size from theprobe tip was measured to be 300 mm FWHM indiameter.

In practice, a single macropulse energy of morethan 10 mJ ~6 3 105 W peak power! was readilyavailable, whereas the surgical requirements calledfor 2–3 mJ per macropulse delivered to the tissue.

Fig. 3. Schematic of the FEL HGW delivery system. The FEL bthrough a 150-mm focal-length lens, and through a 500-mm-diamets within a surgical probe and is protected with a CaF2 window–lenip and circulated through the hollow waveguide.

Fig. 4. Open-sky retinal cutting procedure with HGW-deliveredFEL surgical probe. The arrows show the ablated line on theretina.

At this low energy level, the fluence of the FEL pulsewas equivalent to 2.8–4.2 Jycm2.

Human cadaver eyes were used for open-sky reti-notomies. The threshold for cutting was at a fluenceof approximately 1.8 Jycm2 at 6.45 mm., whereas apractical rate of tissue removal with an acceptableoptical cavitation effect was achieved at a fluence of2.0–3.5 Jycm2. Higher fluence caused vaporizationwith adverse mechanical effects. The histologic re-sult of a lased retinal incision is shown in Fig. 5. Nocollateral damage was observed. This in vitro reti-nal ablation result is consistent with our previous

passes through an adjustable diaphragm to attenuate the beam,hole to couple the beam to the hollow waveguide. The waveguide

the tip. Nitrogen gas was flowed into the Teflon tube to the probe

Fig. 5. Histologic results showing the ablated incision in the ret-ina. No collateral damage was found.

eamer pins at

1 February 2001 y Vol. 40, No. 4 y APPLIED OPTICS 585

iFe

oo

tl

o~1RTAoaJn

5

tissue-ablation experiments.3 This system also hasbeen applied successfully in live-animal surgical pro-cedures11,12 ~Fig. 6!.

4. Discussion

An advantage of the FEL is that it can be tuned to aspecific absorption band of tissue or, more specifi-cally, to a major component of that tissue, in themid-infrared spectrum range. However, the mid-infrared range presents technological challenges formedical-beam delivery. Our experiments haveshown that use of a HGW is a feasible solution.

The flexibility of the HGW is similar to that of asilica glass fiber with the same outer glass diameter.To further protect the HGW used in the experiments,we shielded it with an outer sterilized Teflon tube.The SMA connector permitted ease in alignment andconnection to the FEL.

The 20% injection loss of the waveguide was causedmainly by the large incident angle of the FEL beamand by blocking of the peripheral beam by the 500-mm-diameter pinhole. Because the 3° acceptanceangle of the waveguide was so small, the focusinglens had to be as long as possible while the focal pointfit in the diameter of the waveguide was maintained.Our focused beam had an incident angle of 3.8°,which was slightly larger than the maximum accep-tance angle of the waveguide. The focused spot sizewas 550 mm, which was also slightly larger than thepinhole. When the diaphragm was constricted intoa smaller diameter, a reduction in injection loss wasobserved.

The energy transmission of the waveguide de-pended on the wavelength, the bending curvature,and the diameter of the waveguide. As long as thewaveguide could deliver enough energy to the probetip, the transmission loss would not have any effecton the surgical applications. The change in thetransmission with changes in the bending curvaturemay influence the stability of the delivered laser en-ergy when the surgeon is manipulating the surgicalprobe. Our experiments had shown that in condi-tions of extreme bending curvature the energy vari-

Fig. 6. Surgeons using a HGW-delivered FEL surgical probe toconduct an animal experiment.

86 APPLIED OPTICS y Vol. 40, No. 4 y 1 February 2001

ation was less than 15%, which would not have asignificant effect on the cutting results.

Although we did not conduct damage-thresholdtesting, our experiments suggested that the 530-mmnner-diameter HGW can tolerate more than 24 mJ ofEL energy at the 6.45-mm wavelength, which isquivalent to 7.2 3 108 Wycm2 power density. The

most vulnerable damaged point was the injection endof the waveguide, especially during the alignmentprocedure. If the FEL beam was focused on the edgeof the silver coating or the glass wall, plasma formedand damaged the waveguide end. After the 500-mm-diameter metal pinhole was placed in front of theinjection end of the hollow waveguide, this damagewas avoided.

The delivery system with the hollow waveguideprovides the surgeons with ~1! effective coupling ofthe FEL beam, ~2! flexibility and relative ease of usein microsurgical procedures, ~3! ease of alignmentand connection to the laser, ~4! ability for near contactr contact with tissue and water, and ~5! achievementf sterility with gas sterilization and a disposable tip.These results show that the HGW’s delivery sys-

em is a promising technology for medical beam de-ivery of infrared FEL’s for microsurgical procedures.

This study was supported by the U.S. Departmentf Defense Medical Free Electron Laser ProgramU.S. Office of Naval Research grant N00014-94-023!, the Middle Tennessee Lions Eye Bank, andesearch to Prevent Blindness, Inc. ~New York!.his research was presented in part as a poster at thessociation for Research in Vision and Ophthalmol-gy Annual Meeting, Sarasota, Florida, May 1997,nd at the BIOS’98 meeting, San Jose, California,anuary 1998. The authors acknowledge the tech-ical assistance of Richard Robinson.

References1. C. A. Brau, “The Vanderbilt University free-electron laser cen-

ter,” Nucl. Instrum. Meth. A 319, 47–50 ~1992!.2. G. S. Edwards, R. Logan, M. Copeland, L. Reinisch, J. David-

son, B. Johnson, R. Maciunas, M. Mendenhall, R. Ossoff, J.Tribble, J. Werkhaven, and D. O’Day, “Tissue ablation by afree-electron laser tuned to the Amide II band,” Nature 371,416–419 ~1994!.

3. K. M. Joos, J. H. Shen, G. S. Edwards, D. J. Shetlar, R. D.Robinson, and D. O’Day, “Infrared free electron laser-tissueinteractions with human ocular tissues,” Invest. Ophthalmol.Vis. Sci. Suppl. 37, S431 ~1996!.

4. G. N. Merberg, “Current status of infrared fiber optics formedical laser power delivery,” Lasers Surg. Med. 13, 572–576~1993!.

5. Y. Matsuura, T. Abel, and J. A. Harrington, “Optical propertiesof small-bore hollow glass waveguides,” Appl. Opt. 34, 6842–6847 ~1995!.

6. J. A. Harrington and Y. Matsuura, “Review of hollowwaveguide technology,” in Biomedical Opto-electronic Instru-mentation, A. Katzir, J. A. Harrington, and D. M. Harris, eds.,Proc. SPIE 2396, 4–14 ~1995!.

7. Y. Matsuura, T. Abel, and J. A. Harrington, “Small-bore hollowwaveguide for delivery of near singlemode IR laser radiation,”Electron. Lett. 30, 1688–1690 ~1994!.

8. E. A. J. Marcatili and R. A. Schmeltzer, “Hollow metallic and

dielectric waveguides for long distance optical transmission

1

11. W. Sun, J. H. Shen, D. J. Shetlar, and K. M. Joos, “Endoscopic

1

and lasers,” Bell. Syst. Tech. J. 43, 1783–1809 ~1964!.9. M. Miyagi and S. Karasawa, “Waveguide losses in sharply bent

circular hollow waveguides,” Appl. Opt. 29, 367–370 ~1990!.0. K. Matsuura, Y. Matsuura, and J. A. Harrington, “Evaluation

of gold, silver, and dielectric-coated hollow glass waveguides,”Opt. Eng. 35, 3418–3421 ~1996!.

goniotomy with the free electron laser in congenital glaucomarabbits,” J. Glaucoma ~to be published!.

2. K. Joos, J. Shen, D. Shetlar, and V. Casagrande, “Opticnerve sheath fenestration with a novel wavelength producedby the free electron laser ~FEL!,” Lasers Surg. Med. ~to bepublished!.

1 February 2001 y Vol. 40, No. 4 y APPLIED OPTICS 587