Generalization of Deep Neural Networks for Chest Pathology Classification in X-Rays Using Generative Adversarial Networks

TL;DR

This paper demonstrates that using GAN-generated images to augment imbalanced chest X-ray datasets improves deep neural network classification accuracy for various pathologies.

Contribution

It introduces a method to generate artificial chest X-ray images with GANs to balance datasets and enhance pathology classification performance.

Findings

GAN augmentation improves classification accuracy.

Balancing dataset classes enhances model performance.

Artificial images help detect rare conditions better.

Abstract

Medical datasets are often highly imbalanced with over-representation of common medical problems and a paucity of data from rare conditions. We propose simulation of pathology in images to overcome the above limitations. Using chest X-rays as a model medical image, we implement a generative adversarial network (GAN) to create artificial images based upon a modest sized labeled dataset. We employ a combination of real and artificial images to train a deep convolutional neural network (DCNN) to detect pathology across five classes of chest X-rays. Furthermore, we demonstrate that augmenting the original imbalanced dataset with GAN generated images improves performance of chest pathology classification using the proposed DCNN in comparison to the same DCNN trained with the original dataset alone. This improved performance is largely attributed to balancing of the dataset using GAN generated images, where image classes that are lacking in example images are preferentially augmented.