EyeLoveGAN: Exploiting domain-shifts to boost network learning with cycleGANs
This work addresses medical image analysis for glaucoma diagnosis, but it is incremental as it applies existing methods to a new dataset with a domain adaptation technique.
The authors tackled the REFUGE challenge tasks of retinal image segmentation, classification, and localization by using CNNs, U-Net, InceptionV3, and stacked hour-glass networks, and employed cycleGANs to create domain-shifts between data sources for enhanced performance, though no concrete numbers are provided.
This paper presents our contribution to the REFUGE challenge 2020. The challenge consisted of three tasks based on a dataset of retinal images: Segmentation of optic disc and cup, classification of glaucoma, and localization of fovea. We propose employing convolutional neural networks for all three tasks. Segmentation is performed using a U-Net, classification is performed by a pre-trained InceptionV3 network, and fovea detection is performed by employing stacked hour-glass for heatmap prediction. The challenge dataset contains images from three different data sources. To enhance performance, cycleGANs were utilized to create a domain-shift between the data sources. These cycleGANs move images across domains, thus creating artificial images which can be used for training.