Non uniform fast fourier transform technique to estimate the Stokes spectra for a fiber optic spectral polarimetry instrument

Optical Coherence Tomography (OCT) is an interferometric, noninvasive and noncontact imaging technology that can image biological tissues at micrometer-scale resolutions [1]. OCT detects the interference signal between back-reflected probe light and reference light passing through an interferometer. Unlike sound waves, light as a transverse wave has a polarization property that can be used to enhance contrast via Polarization Sensitive OCT (PS-OCT). The polarization state of a light field can always be completely characterized by a Stokes Vector. In the case of a fiber-optic spectral polarimetry instrument [2], output intensity is a superposition of the four Stokes spectra modulated at different carrier frequencies dependent on phase retardations in polarization maintaining (PM) fiber segments. A simple Fourier transform of the output intensity isolates each Stokes spectral component in the time-delay or optical path length difference domain. The carrier frequency of each Stokes spectral component scales with the length of PM fiber segments and if their lengths are too short, the Stokes parameter components with different carrier frequencies overlap in the optical path length difference domain and cannot be demodulated. During the course of this study, several approaches were studied to tackle the problem of demodulating the Stokes parameters when interfering with each other. The results of the experiments performed using the various methods presented in chapter 4 suggest that the Non uniform fast Fourier transform provides the best result and shows most promise for estimating the Stokes spectra of FOSPI data