Determination of tissue optical properties from spatially resolved relative diffuse reflectance

Noninvasive determination of mu(s)' and mu(a) is essential for clinical applications in medical diagnostics and therapeutics. Spatially resolved diffuse reflectance method is more advantageous than other techniques because of its simplicity and low-cost. The methods for solving the nonlinear inverse problem of estimates of mu(s)' and mu(a) from spatially resolved diffuse reflectance R-d(r) can be classified into the algorithms based on absolute or relative reflectance measurements in nature. Since absolute reflectance measurements are technically more difficult to perform than the relative one, study on the methods based on the relative reflectance has a more important meaning for real applications. Considering that there were several normalizations of R-d(r), in this paper we discussed the varieties of prediction rms errors of mu(s)' and mu(a) extracted from relative reflectance data of different normalization forms including R-d(r)/R-d(r)(max), r(2)(R-d(r)/R-d(r)(max)), ln(R-d(r)/R-d(r)(max)) and ln(r(2)(R-d(r)/R-d(r)(max))). Additionally, we compared the accuracies of mu(s)' and mu(a) determined from absolute reflectance data R-d(r) and ln(R-d(r)) with that from relative reflectance data to study the loss of accuracy due to normalization. Rather than the traditional neural network methods, we used a new method - PCA-NN trained with diffuse reflectance data from Monte Carlo simulations to derive mu(s)' and mu(a). All the PCA-NNs were trained and tested on the space with mu(s)' between 0. 1 and 2.0 mm(-1) and mu(a) between 0.01 and 0.1 mm(-1). The test results indicate that the rms errors in mu(s)' and mu(a) are 0.72% and 2.57% for R-d(r), 0.28% and 0.55% for ln(R-d(r)), 2.98% and 5.44% for R-d(r)/R-d(r)(max), 2.22% and 3.21% for ln(R-d(r)/R-d(r)(max)), 6.52% and 20.7% for r(2)(R-d(r)/R-d(r)(max)), and 2.22% and 3.21% for ln(r(2)(R-d(r)/R-d(r) max)), Suggesting that the normalization form ln(R-d(r)/R-d(r)(max)) would be the first choice for the estimates of As' and A. from relative reflectance data by PCA-NN. Although the loss of accuracy due to normalization is considerable, the preliminary results provide a guideline for relative reflectance measurements.