Visible Reflectance Hyperspectral Imaging: Characterization of a Noninvasive, in Vivo System for Determining Tissue Perfusion Karel J. Zuzak, Michael D. Schaeberle, E. Neil Lewis, ² and Ira W. Levin* Laboratory of Chemical Physics, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland 20892 We characterize a visible reflectance hyperspectral imag- ing system for noninvasive, in vivo, quantitative analysis of human tissue in a clinical environment. The subject area is illuminated with a quartz-tungsten-halogen light source, and the reflected light is spectrally discriminated by a liquid crystal tunable filter (LCTF) and imaged onto a silicon charge-coupled device detector. The LCTF is continuously tunable within its useful visible spectral range (5 2 5 -7 2 5 nm) with an average spectral full width at half-height bandwidth of 0 .3 8 nm and an average transmittance of 10.0%. A standard resolution target placed 5.5 ft from the system results in a field of view with a 1 7 -cm diameter and an optimal spatial resolution of 0.45 mm. The measured reflectance spectra are quantified in terms of apparent absorbance and formatted as a hyperspectral image cube. As a clinical example, we examine a model of vascular dysfunction involving both ischemia and reactive hyperemia during tissue reper- fusion. In this model, spectral images, based upon oxy- hemoglobin and deoxyhemoblobin signals in the 5 2 5 - 6 4 5 -nm region, are deconvoluted using a multivariate least-squares regression analysis to visualize the spatial distribution of the percentages of oxyhemoglobin and deoxyhemoglobin in specific skin tissue areas. Oximetry, an analytical approach for assessing blood flow through the vascular bed of a tissue, monitors the visible and near- infrared spectral properties of blood by recording variations in the percentages of oxygen saturation of hemoglobin. 1-7 The detection of oxy- (HbO 2 ) and deoxyhemoglobin (deoxy-Hb) through the skin by a device consisting of a lamp, filter assembly, and photocell led to the original ear oximeter in which short- and long-wavelength light were passed through the tissue and com- pared. 8 Later, measurements of the oxygen saturation of hemo- globin were refined by comparing the optical densities of blood at 660 nm and at a reference wavelength, 805 nm, the isobestic point for oxy- and deoxyhemoglobin. 9 Additional approaches developed over the years resulted in a clinical oximeter clipped to a patient’s fingertip. 10-11 Although oximetry represents a commonly used technique for determining spectrometric changes in hemoglobin characteristics, the method only measures oxygen saturation at either one or several sample points, which dramatically limits the technique to an assessment of a small region of tissue. Recent oximetry studies using specifically the near-infrared (NIR) spectral region, however, have been used to image tissue perfusion. 12-14 Since the NIR spectral interval is complicated by the presence of both over- lapping water bands and vibrational overtones, oximetry meth- odologies employing visible reflectance data provide a more straightforward manner in, for example, the determination of tissue perfusion and reperfusion as related to clinical events. In addition, oxygen saturation measurements in the clinical environ- ment using visible reflectance imaging, as emphasized in this report, have the distinct advantage of visualizing the capillary circulation processes where oxygen is exchanged, which in this case is within a few millimeters of the skin’s surface. In contrast, NIR spectroscopic and other analytically based phlebotomy techniques typically examine blood characteristics within the deeper and larger blood vessels from which only inferences regarding oxygen exchange are determined. Since the clinical usage of visible reflectance hyperspectral imaging is nonexistent at the present, we developed a technique to collect spectroscopic data from a large array detector at * Corresponding author: National Institutes of Health, Building 5, Room B1- 32, Bethesda MD 20892-0510. (E-mail) iwl@ helix.nih.gov. (Fax) 301-496-0825. † Present address: Spectral Dimensions, Inc., 3416 Olandwood Court, Suite #210 Olney, MD 20832. (1) Cohen, A.; Wadsworth, N. Med. Biol. Eng. 1972 , 10, 385-390. (2) Takatani, S.; Cheung, P. W.; Ernst, E. A. Ann. Biomed. Eng. 1980 , 8,1-15. (3) Hajizadeh-Saffar, M.; Feather, J. W.; Dawson, J. B. Phys. Med. Biol. 1990 , 35, 1301-1315. (4) Marchesini, R.; Brambilla, M.; Clemente, C.; Maniezzo, M.; Sichirollo, A. E.; Testori, A.; Venturoli, D. R.; Cascinelli, N. Photochem. Photobiol. 1991 , 53, 77-84. (5) Malonek, D.; Grinvald, A. Science 1996 , 272, 551-554. (6) Knoefel, W. T.; Kollias, N.; Rattner, D. W.; Nishioka, N. S.; Warshaw, A. L. J. Appl. Physiol. 1996 , 80, 116-123. (7) Malonek, D.; Dirnagl, U.; Lindauer, U.; Yamada, K.; Kanno, I.; Grinvald, A. Proc. Natl. Acad. Sci. U.S.A. 1997 , 94, 14826-14831. (8) Millikan, G. A. Rev. Sci. Instrum. 1942 , 13, 434-444. (9) Wood, E. H.; Geraci, J. E. J. Lab. Clin. Med. 1949 , 34, 387-401. (10) Gordy, E.; Drabkin, D. C. J. Biol. Chem. 1957 , 227, 285-299. (11) Yoshiya, I.; Shimada, Y.; Tanaka, K. Med. Biol. Eng. Comput. 1980 , 18, 27-32. (12) Stranc, M. F.; Sowa, M. G.; Abdulrauf, B.; Mantsch, H. H. Br. J. Plast. Surg. 1998 , 51, 210-217. (13) Sowa, M. G.; Payette, J. R.; Mantsch, H. H. J. Surg. Res. 1999 , 86, 62-69. (14) Sowa, M. G.; Leonardi, L.; Payette, J. R.; Fish, J. S.; Mantsch, H. H. Burns 2001 , 27, 241-249. Anal. Chem. 2002, 74, 2021-2028 10.1021/ac011275f Not subject to U.S. Copyright. Publ. 2002 Am. Chem. Soc. Analytical Chemistry, Vol. 74, No. 9, May 1, 2002 2021 Published on Web 04/05/2002