【正文】 l 703T) emitting radiation that was coupled into a 600 core diameter optical ?ber (Thor Labs, Newton, NJ). A steppermotordriven translation stage (Newport, Irvine, CA) scanned the laser beam along the axis of the weld site at speeds that effectively produced 100mslong pulses. Seventy scans were made along each weld。 the beam stopped at the end of the weld site for 10 s after each scan. To minimize thermal damage to the skin beyond the weld area, highre?ecting metal plates placed on each end of the incision blocked the beam. Experiments were performed at constant irradiance (127) paring laser spot diameters of 1, 2, 4, and 6 mm [fullwidth at fullmaximum(FWHM)], with laser output powers of 1, 4, 16, and 36 W, respectively. The beam pro?le, as measured by scanning a 200 mdiameter pin hole across the beam, was approximately Gaussian for all spot diameters. The power delivered to the tissue was measured before each weld with a power meter (Molectron PowerMax 5100, Portland, OR). It shows the experimental con?guration used for dyeassisted laser skin welding and summarizes the laser parameters for this study.After welding, the anesthetized guinea pig was euthanized with an intracardiac overdose of sodium pentobarbitol (Nembutal, Abbott Laboratories, North Chicago, IL). The dorsal skin, including epidermis and dermis, was excised with a scalpel and then sectioned. Samples were processed using standard histological techniques, including storage in 10% formalin, processing with graded alcohols and xylenes, para?n embedding, sectioning, and hemotoxylin and eosin staining. A minimum of seven samples was processed for each laser spot diameter and beam pro?le. The 6mmdiameter spot study was discontinued after grossly obvious burns developed at the wound site.Thermal denaturation measurements were made using a transmission light microscope (Nikon, Japan) ?t with crossed linear polarizers (Prinz, Japan). Thermal denaturation was measured laterally from the center of the weld site at three different depths: the papillary dermis, middermis, and base of the dermis. The depth to which one observed denaturation was recorded and divided by the skin thickness to obtain the fraction of a fullthickness weld that was achieved. Measurements were made consistently to the point at which plete thermal denaturation of the tissue was observed.Statistical analyzes were conducted on the histological data. ANOVA was used to determine statistical signi?cance of thermal denaturation measurements between laser spot size groups.B. Monte Carlo SimulationMonte Carlo simulations were run to investigate the effect of various spot sizes (1–6mm diameters) and beam pro?les (Gaussian versus ?attop and single versus dual beam) on the distribution of absorbed radiation. All simulations were run using code available over the public domain . Several changes were made in the Monte Carlo code to adapt it for use with the geometry of this application. First, because the vertical ink layer in the tissue disrupted the cylindrical symmetry assumed in the Original program, the data were stored in Cartesian rather than cylindrical coordinates and a convolution program was not used to generate the laser beam pro?le. The beam pro?le was, instead, created using a random number generator 。 a large number of photons was used to create the desired beam pro?le. Second, the vertical ink layer was modeled as an in?nite absorber extending from the skin surface to the base of the dermis with a uniform thickness of 100 m. The experimentally measured absorption coef?cient for the ink, was 3500 cm. Even though histologic analysis of the welds showed variable staining of the tissue with a lateral thickness varying from 40 to 100 m, since the ink layer thickness was much greater than the probability that a photon could cross the ink layer was negligible, and the assumption that was in?nite is reasonable. Third, the skin was modeled as a single dermal tissue layer with the assumption that the epidermis and subcutaneous tissue have optical properties similar to that of the dermis. Finally, even though the optical properties of tissue are known to be temperaturedependent, with the dermal scattering coef?cient initially increasing with temperature for temperatures less than 60 C then decreasing sharply at higher temperatures and the dermal absorption coef?cient decreasing with increasing temperature , the optical properties in this model were assumed to be static. This assumption, which avoided a plete opticalthermal model, will result in a slight underestimation of the penetration depth of the photons in the dermis. The optical properties of guinea pig skin at a wavelength of have not been well characterized. The optical properties for human, pig, and rat dermis were therefore. piled from several sources. The optical properties used in the Monte Carlo simulations are listed in Table II. Note that in the experimental irradiations, the irradiance was held constant at 127 . For the simulated irradiations, the mean irradiance over the fullwidth, halfmaximum of each beam was constant (10 photons per 1mmdiameter area). The grid element size in the tissue was ?xed at 100 m, and the dimensions of the tissue (length width depth) were cm cm cm, respectively. The tissue thickness was, in part, chosen based on the knowledge that human skin may be thicker than guinea pig skin, ranging in thickness from 1 to 4 mm. Simulations were run on a Pentium 133 MHz PC puter (Micron, Nampa, ID)running Microsoft Windows 95 (Microsoft, Redmond, WA)III. RESULTSA. ExperimentsHistologic analysis showed that only shallow welds were achieved using a 1mmdiameter laser irradiation area. Thermal denaturation was observed only to a depth of 570100 (.,n=7) or 30% of the average dermal thickness of 1900200 , see Table III. Thermal denaturation lateral to the incision was limited to near the tissue surface. An image of a weld created with a 1mmdiameter s