Quantitative Optical Coherence Elastography: A Novel Intensity-Based Inversion Method Versus Strain-Based Reconstructions

Abstract

Many diseases like cancer or atherosclerosis bear micro-scale tissue stiffness changes, which can be visualised by optical coherence elastography (OCE). OCE is a promising research field, where a growing number of publications present qualitative displacement maps. Quantitative OCE results have also been presented, but still lack high precision and good reproducibility, which are important for clinical applications. In this work, we compare three reconstruction methods for the Young's modulus in intensity-based quasi-static compressional OCE: uniaxial analysis, strain map based reconstruction facilitating a particle tracking improved optical flow (EOFM), and a novel image-based inverse reconstruction method (IIM). The quality of the proposed reconstruction methods is investigated by comparing their performance on twelve silicone elastomer phantoms with inclusions of varying size and stiffness. While the uniaxial reconstruction is strongly affected by lateral motion, EOFM is capable of deriving precise strain maps from consecutive OCT images during compression. However, for a valid Young's modulus reconstruction additional stress map information is required. IIM performs best, precisely reconstructing Young's modulus of inclusion and background, closely corresponding to separately determined ground truth values. The method offers a significant improvement of the relative error by a factor of 3.5 compared to uniaxial and strain-based analysis.