An Evaluation of the Nonlinear MultiComponent Analysis Technique for Reflection Spectra

Analysis of the absorption spectra in the visible range has made it possible to monitor the oxygen supply and metabolism of organic tissues continuously and non-invasively. The nonlinear multicomponent analysis (NLMCA) is an evaluation method for the reflection spectra of the tissues. In this report, the derivation of the relation between the reflection and absorption spectra based upon Kubelka-Monk theory is reviewed to introduce the foundation of the NLMCA technique. Then, the procedures of the NLMCA algorithm are described and simulation studies are performed. The results indicate that the NLCMA algorithm generally works well after a minor modification but is sensitive to modelling errors due to invalid assumptions on the scattering properties of samples. A direct algorithm based on linear least squares is the suggested to deal with the same problem and simulation studies are performed to make a comparison with the NLMCA algorithm, which demonstrate a better robust properties of the alternative strategy.