Toward absolute reflectance oximetry: I. Theoretical consideration for noninvasive tissue reflectance oximetry.

The photon diffusion theory can yield quantitative estimation of tissue hemoglobin saturation, provided that the medium is homogeneous and that one calibration data is available. The error in detection of tissue OS of the gut mucosa ranged from 5 to 10% in oxygen saturation. In application to skin, the two-layer tissue model suggests that by properly designing the optical sensor and by appropriately selecting the illumination wavelengths, it is possible to capture mainly the light returning from the specific depth in tissue. Since the skin layer thickness is roughly in the order of 1 mm, the source and detector separation distance of approximately 3 mm or larger would ensure that the measured reflectance is truly returning from the deeper layer. When such reflectances are normalized to the blood-free reflectance obtained by squeezing the blood out of the tissue, the normalized reflectance truly represents the deeper layer characteristics. In application to head, since the skin and skull thickness is considerable large, separation distance of 40 mm or greater is required to ensure the reflectance is actually returning from the brain. Closely spaced optical sensor would measure the scattering and absorption characteristics of the skin and skull of the head. As for directional changes in optical propagation due to tissue inhomogeneities, multiple light sources at the equi-distance around the detector can be placed to average out the effect. The resultant reflectance can be analyzed based on the similar mathematical treatment as presented in this study. However, since the absolute reflectance level calculated by the theory and the actual reflectance for a given transducer geometry have some deviation, again one point calibration is required to close the gap between them. This can be accomplished through arterialization of the tissue and ventilating with pure oxygen to yield reflectance from tissue containing 100% saturated blood. As for hemoglobin content, isosbestic reflectance, for example at 805 nm, can be utilized to estimate tissue hemoglobin content. Once one point calibration is accomplished, reflectance changes thereafter due to changes in HbT and OST can be fairly accurately predicted by the photon diffusion theory in combination with linear analysis. Concerning separation of arterial and venous blood in tissue, the diastolic and systolic phases of the optical plethysmographic signal can be assumed to relate to venous or DC level, and to arterial or AC component. Since the four components, arterial and venous OS and Hb, are unknowns in the system, four equations or four wavelength measurements are required to sort out each effect.(ABSTRACT TRUNCATED AT 400 WORDS)