On Urinary Bladder Cancer Diagnosis: Utilization of Deep Convolutional Generative Adversarial Networks for Data Augmentation

Simple Summary One of the main challenges in the application of Machine Learning in medicine is data collection. Either due to ethical concerns or lack of patients, data may be scarce. In this paper Deep Convolutional Generative Adversarial Networks (DCGAN) have been applied for the purpose of data augmentation. Images of bladder mucosa are used in order to generate new images using DCGANs. Then, combination of original and generated images are used to train AlexNet and VGG16 architectures. The results show improvements in AUC score in some cases, or equal scores with apparent lowering of standard deviation across data folds during cross-validation; indicating networks trained with the addition of generated data have a lower sensitivity across the hyperparameter range. Urinary bladder cancer is one of the most common urinary tract cancers. Standard diagnosis procedure can be invasive and time-consuming. For these reasons, procedure called optical biopsy is introduced. This procedure allows in-vivo evaluation of bladder mucosa without the need for biopsy. Although less invasive and faster, accuracy is often lower. For this reason, machine learning (ML) algorithms are used to increase its accuracy. The issue with ML algorithms is their sensitivity to the amount of input data. In medicine, collection can be time-consuming due to a potentially low number of patients. For these reasons, data augmentation is performed, usually through a series of geometric variations of original images. While such images improve classification performance, the number of new data points and the insight they provide is limited. These issues are a motivation for the application of novel augmentation methods. Authors demonstrate the use of Deep Convolutional Generative Adversarial Networks (DCGAN) for the generation of images. Augmented datasets used for training of commonly used Convolutional Neural Network-based (CNN) architectures (AlexNet and VGG-16) show a significcan performance increase for AlexNet, where AUCmicro reaches values up to 0.99. Average and median results of networks used in grid-search increases. These results point towards the conclusion that GAN-based augmentation has decreased the networks sensitivity to hyperparemeter change.