Study on Internal Information of the Two-Layered Tissue by Optimizing the Detection Position

As to most methods of detecting the inner information of inhomogenous tissue, a significant issue is that the detection position is ambiguous because of the complexity of human tissue structure and discrepancies among individuals. This paper studies the best source-detector distance (SDSbest) to detect internal information of a fat-muscle tissue with spatially resolved diffuse reflectance spectra. In order to weaken the measurement error caused by the discrepancies among individuals and multiple backscattered photons, and according to the transmission model of light in complex biological tissue, then we added the constraint condition two ideal "banana shape" paths to define the effective photon ratio(SNR), which was used to select the best source detector separations (SDSbest), and the results from Monte Carlo simulation modified by adding constraint condition were statistically analyzed, and we regard the SNR as a basis and analyze the relationship between the fat thickness (h(f)), the absorption coefficient of a fat layer (mu(af)), the absorption coefficient of a muscle layer (mu(am)) and the source-detector distance (SDS), and h(f) is used as the independent variable to develop a linear regression model to predict SDSbest. The result showed that mu(af) and mu(am), have no effect on mu(af) when 0<h(f) <0. 6 cm, and the correlation coefficient of the linear regression model is 0. 991 8; Randomly select h(f) = 0. 12 and 0. 22 cm, the prediction error is 0. 030 14 and 0. 020 16 respectively, the error can be controlled within 5%. This method can select the SDSbest much easier and faster to detect the inner information of turbid tissue, and to weaken the interference from the non-target layer and multiple backscattered photons.