A method of complex conjugate artifact elimination for SS-OCT systems based on adversarial generative model

Swept source optical coherence tomography is a common medical instrument used for ophthalmic clinical examination. However, the presence of complex conjugate artifacts may cause severe occlusion of eye cross-sectional structure in B-scan images. Existing high-speed complex conjugate artifact elimination methods often require additional hardware in the system, increasing the system complexity and manufacturing cost. In this article, we propose a two-step complex conjugate artifact elimination method. Firstly, we detect the position of complex conjugate artifacts morphological image processing methods. Subsequently, a proposed adversarial generative model is used to translate images with complex conjugate artifacts into images without complex conjugate artifacts. We trained and tested the model on a self-built dataset to verify the feasibility of this method. Quantitative and qualitative analysis of the results show that this method can effectively suppress complex conjugation artifacts and preserve the correct artifact-free fundus structure.