Impact of multiplicative noise removal on digital volume correlation-based optical coherence elastography

Digital volume correlation-based optical coherence elastography (DVC-based OCE) may easily suffer from multiplicative speckle noise of optical coherence tomography (OCT) imaging, which can introduce non-negligible calculation errors. In this study, Hamilton-Jacobi partial differential equations were adopted for multiplicative noise removal of OCT images and the impact of multiplicative noise removal on DVC-based OCE was investigated. Several deformation conditions, including static, sub-pixel translation, uniform compression and non-uniform deformation, were tested. Results showed that multiplicative noise removal can suppress the maximum noise-induced error to less than 0.15 pixels in static tests. A more precise sub-pixel translation with an accuracy of 0.6 pixels was obtained. For uniform compression, multiplicative noise removal can extend the strain upper limit to 0.116 with an error less than 15% and a correlation value higher than 0.8. It also worked better in non-uniform deformation with more reliable calculation points. These findings provide a way to improve the measurement accuracy of the DVC-based OCE method by removing multiplicative noise on OCT images.