Generative adversarial network-convolution neural network based breast cancer classification using optical coherence tomographic images

Abstract

Currently, breast tissue images are primarily classified by pathologists, which is time-consuming and subjective. Deep learning, however, can perform this task with the utmost precision. In order to achieve an improved performance, a large number of annotated datasets are required to train the network, which is a challenging task in the medical field. In this paper, we propose an intelligent system, based on generative adversarial networks (GANs) and a convolution neural network (CNN) for the automatic classification of breast cancer, using optical coherence tomography (OCT) images. In this network, the GAN is used to generate synthetic datasets and to further utilize these synthetic datasets to increase the quantity of information, so as to improve the classification performance of the CNN. Our method is demonstrated by means of a limited set of OCT images of breast tissue. The classification performance of our method, using only the classic data increase, yielded a sensitivity level of 93.6%, with 90.8% specificity and 91.7% accuracy, based on the test datasets. By adding the synthetic data increase, the accuracy of the training datasets increased to 93.7% from 92.0%. We believe that this approach will help radiologists and pathologists to improve their diagnotic capability.