A two-step iteration mechanism for speckle reduction in optical coherence tomography

• Main Authors: Wang, Xianghong, Yu, Xiaojun, Liu, Xinyu, Chen, Si, Chen, Shufen, Wang, Nanshuo, Liu, Linbo
• Other Authors: School of Electrical and Electronic Engineering
• Format: Article
• Language: English
• Published: 2019
• Subjects: Optical Coherence Tomography; Image Processing; DRNTU::Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/90101; http://hdl.handle.net/10220/48378
• Institution: Nanyang Technological University

Optical coherence tomography (OCT) is an imaging tool that has been widely utilized for various disease diagnoses for its noninvasive and high-resolution properties. Due to the dual role of speckles in the imaging process, however, OCT images suffer from the unavoidable speckle noise, which is usually regarded to be multiplicative in nature and reduces image contrast and resolution. In this study, we propose to categorize OCT speckle noise into additive portion and multiplicative portion, and present a simple two-step iteration (TSI) mechanism to suppress such noises separately. With the augmented Lagrange minimization (ALM) method adopted to recover a low-rank image, the first step of TSI is to remove the additive Gaussian noise, while the second step of TSI is to suppress the multiplicative noise by employing a split Bregman method (SBM) to solve the total-variation (TV) de-noising problem. Extensive experiments with OCT images of the swine eye, human and rabbit retina are conducted to verify the effectiveness of the proposed method. Results show that the proposed TSI method outperforms the existing methods in different cases. Specifically, TSI helps improve the peak signal-to-noise ratio (PSNR) and structure similarity (SSIM) of the swine eye images from 17.19 dB to 33 dB and 0.12 to 0.92, respectively, with the important structural details well preserved. The clearer boundaries and higher image contrast obtained with the TSI method would largely facilitate image analyses and interpretations for the clinical applications of OCT systems.