Collection efficiency of a single optical fiber in turbid media for reflectance spectroscopy Paulo R. Bargo, Scott A. Prahl and Steven L. Jacques Oregon Medical Laser Center, Providence St. Vincent Hospital, 9205 SW Barnes Rd,. Portland, OR OGI School of Science and Engineering, Oregon Health and Science University, 20000 NW Walker Rd., Beaverton, OR 97006 (503) 216-4093, (503)216-2422, pbargo@ece.ogi.edu, prahl@ece.ogi.edu , sjacques@ece.ogi.edu Abstract: The effect of optical properties on the optical fiber collection efficiency in turbid media was studied experimentally and modeled by Monte Carlo simulations. An analytic expression was obtained to estimate the collection efficiency. 2001 Optical Society of America OCIS codes: (170.3660) Light propagation in tissues; (170.3890); Medical optics instrumentation 1. Introduction Single optical fibers have been commonly used as light delivering and collection tools for optical diagnosis. Authors have proposed their use to determine tissue optical properties [1], measure chromophore relative concentration [2-4] and monitor drug pharmacokinetics [5]. This paper explores the effect of the tissue optical properties on the optical fiber collection efficiency (f) when a single optical fiber is used as light source and detector. The optical fiber collection efficiency is defined as the total number of photons collected by the optical fiber that exit the tissue with an angle less than the fiber acceptance angle (numerical aperture – NA [6]) divided by all the photons collected by the fiber. In other words, f is the number of photons that couple to the fiber CORE divided by the number of photons that couple to the fiber CORE and CLAD as stated in eq. 1, and can be determined by Monte Carlo simulations. f CORE CORE CLAD = + (1) f CORE CORE CLAD AIR * = + + (2) It is difficult to determine f experimentally since the light lost in the fiber cladding is difficult to measure accurately. For that reason a different collection efficiency (f*, eq. 2; defined by Saidi [7]) was determined experimentally to compare experiments to theory. The term AIR is used for all the photons that exit the medium outside the optical fiber. The sum CORE + CLAD + AIR is equal to the total diffuse reflectance exiting the tissue and can be calculated by Monte Carlo simulations [8, 9] or Adding Doubling [9]. 2. Materials and Methods 2.1. Acrylamide Gel Optical Phantoms Preparation A 6x2 matrix of different optical property acrylamide gel phantoms was prepared using Intralipid as scattering element and India ink as absorber. Stock Intralipid-20% was calibrated with the added absorber technique. Stock Intralipid-20% (μ s ′ 2245 200 cm -1 at 630nm) was mixed with water and stock India ink (μ a 2245 58 cm -1 at 630nm) in appropiate proportions. Solutions had absorption coefficients (μ a ) of 0.01, 0.1, 0.5, 1, 5 and 10 cm -1 and reduced scattering coefficients (μ s ′) of 10 and 20 cm -1 at 630nm with a final volume of 40ml. Gels were prepared by adding 8g of acrylamide gel crystals, 0.16g of BIS-acrylamide, 0.100g of amonium persulphate and 0.2ml of TEMED to the solutions while stirring at 38°C. Samples gelled after approximate 2 minutes. Final sample volume was 48ml and samples were assumed to be a semi-infinite homogeneous medium for the purpose of modeling. 2.2. Monte Carlo Simulations Monte Carlo simulations [8, 9] were performed for the same set of optical properties to establish f. Photons were randomly launched within the radius of the fiber forming a collimated beam into a homogenous semi-infinite medium. Henyey-Greenstein phase functions were used to randomly assign propagation angles (θ). The average cosine (or anisotropy, g) was fixed to a typical tissue value of 0.9. Photons were tracked until totally absorbed or until crossing the air/medium boundary. The roulette method [8, 9] was used to conserve energy. Photons that crossed the boundary were summed into three different bins according to the exit location at the surface and the exiting propagation angle (outside fiber = AIR, inside the fiber = CORE, for θ < NA and = CLAD for θ > than NA). 2.3. Single fiber Reflectance Measurements Samples were measured with a single 600μm optical fiber (Tefzel, Ceramoptec) coupled to a bifurcated optical bundle through a SMA adapter. The bifurcated bundle was composed of two 300μm optical fibers (Tefzel, P. R. Bargo, S. A. Prahl, S. L. Jacques, “Collection efficiency of a single optical fiber in turbid media for reflectance spectroscopy,” in OSA Biomedical Topical Meetings, pp. 604–606, (2002).