Fenton B M, Gayeski T E
Department of Anesthesiology, University of Rochester Medical Center, New York 14642.
Am J Physiol. 1990 Dec;259(6 Pt 2):H1912-20. doi: 10.1152/ajpheart.1990.259.6.H1912.
Although a four-wavelength method for cryospectrophotometric measurement of intravascular oxyhemoglobin (HbO2) saturations has previously been described, the relationship between experimental measurements and theory has not been clearly detailed. The current work utilizes an empirical relationship between HbO2 saturation measurements and reflected light oximetry, which is consistent with the two-flux theory of Kubelka and Munk (Z. Tech. Phys. 11a: 593-603, 1931). To obtain linear, concentration-independent calibration curves, the theoretical results require that 1) the complex function relating optical density, scattering, and absorption can be linearly approximated over the range of scattering and absorption coefficients used; and 2) the scattering coefficient is independent of wavelength. Incident light cannot easily be measured using reflection spectroscopy, which precludes the determination of isosbestic points. Therefore, equibestic wave-length pairs were used at which optical density differences were invariant with saturation. This allows numerous wavelength sets over the range 540-600 nm to be selected, rather than the limited choices of isosbestic wavelengths. Finally, the effects of freeze rate, freeze depth, Hb concentration, and vessel diameter are each discussed in terms of their influence on experimental measurements.
