Summary

We use a model-based approach to push the resolution of OCT beyond the coherence-length limit. We reformulate the imaging task as a parameter-estimation problem, and derive some corresponding algorithms.

Introduction

Optical Coherence Tomography (OCT) is a recent high-resolution imaging technique based on an interferometric measure in which a wave with low temporal coherence is split into two parts. One of the beams hits the sample, is backscattered, and guided to a photodetector where it interferes with the second beam which acts as a reference wave. By varying the pathlength of the reference wave, one is able to scan inside the object at depths such that the backscattered wave stays coherent with the incident one.

Main Contribution

Retrieving the optical parameters of the object from the OCT measure is an inverse problem for which we have formulated a new exact solution. Our preliminary validation results show that the resolution depth of the reconstructed object depends on the signal-to-noise ratio of the OCT data, and not on the temporal coherence of the illuminating source—contrary to the classical OCT theory.