OCT measures the round trip or double pass transmission through tissue. Estimating the effect on the polarization states of the forward transmission through an optical element based on the measurements of the double pass transmission is not possible in general. However, in the case of tissue, linear birefringence induced by the fibrillar geometry of many tissue building blocks dominates and exceeds the optical activity that may typically be present in biological samples. From the double pass transmission of a homogeneous linearly birefringent (and diattenuating) element, the single pass transmission can be obtained by taking its matrix root. The optic axis of a linear retarder lies in the QU-plane of the Poincaré sphere, and transmission in the reverse direction through the same polarization element is identical to the forward transmission, resulting in a specific matrix symmetry. However, a stack of differently oriented linear retarders corresponds to a general retarder without any symmetry. Yet, by virtue of the imaging geometry, its double pass transmission always appears as a linear element. Reconstructing the depth-resolved optic axis orientation from PS-OCT measurements relies on the assumption that the axial resolution is sufficient to resolve homogenously retarding tissue layers and estimate their forward transmission, one layer at a time [6 (link),7 (link)].
Figure 1 outlines the necessary processing steps to retrieve depth-resolved optic axis orientation from measurements with our intravascular PS-OCT system and explains the catheter geometry. The goal is to reconstruct the optic axis orientation within the plane orthogonal to the imaging beam at each angular position of the fiber probe. Whereas the measurements are performed in a static laboratory frame, the optic axis orientation is expressed within the local coordinate system of the rotating probe beam, corresponding to the longitudinal and circumferential direction of the vessel lumen, respectively. We demonstrate the rich optic axis orientation contrast by processing in vivo measurements of coronary arteries previously acquired during a clinical pilot trial of intravascular polarimetry [11 (link)].