Simultaneous wound border segmentation and tissue classification using a conditional generative adversarial network

Abstract Generative adversarial network (GAN) applications on medical image synthesis have the potential to assist caregivers in deciding a proper chronic wound treatment plan by understanding the border segmentation and the wound tissue classification visually. This study proposes a hybrid wound border segmentation and tissue classification method utilising conditional GAN, which can mimic real data without expert knowledge. We trained the network on chronic wound datasets with different sizes. The performance of the GAN algorithm is evaluated through the mean squared error, Dice coefficient metrics and visual inspection of generated images. This study also analyses the optimum number of training images as well as the number of epochs using GAN for wound border segmentation and tissue classification. The results show that the proposed GAN model performs efficiently for wound border segmentation and tissue classification tasks with a set of 2000 images at 200 epochs.