Evaluation of cartilage properties from NIR diffuse reflectance spectra: numerical simulation

Diseased conditions of articular cartilage such as osteoarthritis (OA) are accompanied by a change in the water concentration, together with a decrease in cartilage thickness. These parameters are important indicators of the normal joint functioning. Light scattering in a heterogeneous joint tissue is a complex process and it is important to supplement its study with understanding gained from modelling. In this work we employed Monte Carlo (MC) for analysis of the relationship between changes in cartilage parameters and calculated diffuse reflection spectra (DRS). It has been shown that water absorption peak at 980 nm is determined by both water content and cartilage thickness, while changes in DRS spectra at 1450 nm are mostly driven only by water content. Since independent assessment of these parameters is clinically relevant, several approaches for their determination were developed. A heuristic algorithm for cartilage thickness determination and its quality metrics (R2= 0.96) were obtained. Several regression models using the features of DRS signal formation at different wavelengths were developed and it was shown that the use of a wide range of wavelengths (850 – 1700 nm) allows to determine the cartilage water concentration and its thickness with the accuracy R2 0.96 and 0.99 while measuring at different probe-detector distances.