Cervical cancer is one of the leading causes of death in women across the globe. Among the several diagnosis techniques that have been explored, optics-based methods have demonstrated highly promising outcomes in detection of cervical cancer at early stages [1]. Mueller matrix imaging and spectroscopy is one of the polarizations based optical technique which has been used in classification of cervical cancer. Polarization parameters obtained from the Mueller matrix have indicated the changes in real and imaginary part of refractive indices between two orthogonal polarizations. In this study, we develop the Fourier domain Mueller matrix imaging (FMMI) system to look at correlations between the real and imaginary parts of the refractive indices. However, the dependence of both parts of refractive indices to one another needs to be established. Simulation studies were performed on 1-D sinusoidal functions and fractional Brownian motion signals as inputs of refractive indices. The correlations between these functions (refractive indices) obtained through Fourier domain Mueller matrix and pairwise linear correlation coefficients were found to be identical. Further, a FMMI system has been developed and calibrated using standard samples (like homogenous diattenuator and linear retarder) with an error <10 %. Currently, experiments are being performed on unstained cervical tissue sections to understand the correlations between refractive indices.
|