Oximetry based on diffuse photon density wave differentials

The quantification of tissue optical properties for calculating blood satur ation and hemoglobin concentration using measurements of diffuse photon den sity waves at some distance away from an intensity-modulated light source, generally requires the determination of the amplitude and phase of this lig ht source. This determination may become a severe impediment for measuremen ts performed in the clinical environment. In this work we extend a self-cal ibrating methodology developed for constant wave and modulation depth-phase measurements, to include amplitude and phase measurements of diffuse photo n density waves. The method uses amplitude and phase changes of intensity m odulated light, under the assumption of known index of refraction and invar iant reduced scattering coefficient mu'(s), to quantify the absorption coef ficient mu(a) without requiring initial amplitude and phase knowledge. Quan tification of the mu(a) at selected time points during a measurement can th en be employed to calibrate numerical solutions of the diffusion equation a nd compute the mu(a) for the remaining time points of the experiment. It is shown that the method is quite insensitive to the knowledge of the exact m u'(s) value so that an assumption on the average mu'(s) value for the tissu e measured may be employed. The sensitivity of calculating blood saturation and hemoglobin concentration, as a function of the deviation of the mu'(s) used in the calculation versus the real mu'(s) value is investigated using simulated data. It is also demonstrated that the saturation calculation is especially insensitive to the mu'(s) guess. The performance of the method to quantify blood oxygen saturation and the concentrations of oxy- and deox y-hemoglobin is examined with experimental measurements at two wavelengths on specially constructed blood model phantoms. To validate the method the m easurements are monitored by a time-resolved spectrometer. The method is sh own to be accurate to within +/-5% in calculating blood saturation and to w ithin +/-10% in calculating hemoglobin concentration compared to the result s obtained with the time-resolved spectrometer and the expected theoretical values. (C) 2000 American Association of Physicists in Medicine. [S0094-24 05(00)00302-3].