Multiple polynomial regression method for determination of biomedical optical properties from integrating sphere measurements

We present a new, to our knowledge, method for extracting optical propertie s from integrating sphere measurements on thin biological samples. The meth od is based on multivariate calibration techniques involving Monte Carlo si mulations, multiple polynomial regression, and a Newton-Raphson algorithm f or solving nonlinear equation systems. Prediction tests with simulated data showed that the mean relative prediction error of the absorption and the r educed scattering coefficients within typical biological ranges were less t han 0.3%. Similar teats with data from integrating sphere measurements on 2 0 dye-polystyrene microsphere phantoms led to mean errors less than 1.7% be tween predicted and theoretically calculated values. Comparisons showed tha t our method was more robust and typically 5-10 times as fast and accurate as two other established methods, i.e., the inverse adding-doubling method and the Monte Carlo spline interpolation method. (C) 2000 Optical Society o f America.