The nonlinear effect of SNR to phase retardation and birefringence measurement was previously formulated in detail for a Jones matrix OCT (JM-OCT) . Based on this, we had developed a maximum a-posteriori (MAP) estimator and quantitative birefringence imaging was demonstrated . However, this first version of estimator had a theoretical shortcoming. It did not take into account the stochastic nature of SNR of OCT signal.
In this paper, we present an improved version of the MAP estimator which takes into account the stochastic property of SNR. This estimator uses a probability distribution function (PDF) of true local retardation, which is proportional to birefringence, under a specific set of measurements of the birefringence and SNR. The PDF was pre-computed by a Monte-Carlo (MC) simulation based on the mathematical model of JM-OCT before the measurement. A comparison between this new MAP estimator, our previous MAP estimator , and the standard mean estimator is presented. The comparisons are performed both by numerical simulation and in vivo measurements of anterior and posterior eye segment as well as in skin imaging. The new estimator shows superior performance and also shows clearer image contrast.