Data imbalance is a classic problem in image classification, especially for medical images where normal data is muchmore than data with diseases. To make up for the absence of disease images, methods which can generate retinal OCTimages with diseases from normal retinal images are investigated. Conditional GANs (cGAN) have shown significantsuccess in natural images generation, but the applications for medical images are limited. In this work, we propose anend-to-end framework for OCT image generation based on cGAN. The new structural similarity index (SSIM) loss isintroduced so that the model can take the structure-related details into consideration. In experiments, three kinds ofretinal disease images are generated. The generated images assume the natural structure of the retina and thus arevisually appealing. The method is further validated by testing the classification performance trained by the generatedimages.
展开▼
