Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Performance of GAN-based augmentation for deep learning COVID-19 image classification

Oleksandr Fedoruk, Konrad Klimaszewski, Aleksander Ogonowski, Rafał Możdżonek

TL;DR

This work considers the multi-class classification problem of chest X-ray images including the COVID-19 positive class that hasn't been yet thoroughly explored in the literature and finds the GAN augmentation approach to be subpar to classical methods for the considered dataset.

Abstract

The biggest challenge in the application of deep learning to the medical domain is the availability of training data. Data augmentation is a typical methodology used in machine learning when confronted with a limited data set. In a classical approach image transformations i.e. rotations, cropping and brightness changes are used. In this work, a StyleGAN2-ADA model of Generative Adversarial Networks is trained on the limited COVID-19 chest X-ray image set. After assessing the quality of generated images they are used to increase the training data set improving its balance between classes. We consider the multi-class classification problem of chest X-ray images including the COVID-19 positive class that hasn't been yet thoroughly explored in the literature. Results of transfer learning-based classification of COVID-19 chest X-ray images are presented. The performance of several deep convolutional neural network models is compared. The impact on the detection performance of classical image augmentations i.e. rotations, cropping, and brightness changes are studied. Furthermore, classical image augmentation is compared with GAN-based augmentation. The most accurate model is an EfficientNet-B0 with an accuracy of 90.2 percent, trained on a dataset with a simple class balancing. The GAN augmentation approach is found to be subpar to classical methods for the considered dataset.

Performance of GAN-based augmentation for deep learning COVID-19 image classification

TL;DR

This work considers the multi-class classification problem of chest X-ray images including the COVID-19 positive class that hasn't been yet thoroughly explored in the literature and finds the GAN augmentation approach to be subpar to classical methods for the considered dataset.

Abstract

The biggest challenge in the application of deep learning to the medical domain is the availability of training data. Data augmentation is a typical methodology used in machine learning when confronted with a limited data set. In a classical approach image transformations i.e. rotations, cropping and brightness changes are used. In this work, a StyleGAN2-ADA model of Generative Adversarial Networks is trained on the limited COVID-19 chest X-ray image set. After assessing the quality of generated images they are used to increase the training data set improving its balance between classes. We consider the multi-class classification problem of chest X-ray images including the COVID-19 positive class that hasn't been yet thoroughly explored in the literature. Results of transfer learning-based classification of COVID-19 chest X-ray images are presented. The performance of several deep convolutional neural network models is compared. The impact on the detection performance of classical image augmentations i.e. rotations, cropping, and brightness changes are studied. Furthermore, classical image augmentation is compared with GAN-based augmentation. The most accurate model is an EfficientNet-B0 with an accuracy of 90.2 percent, trained on a dataset with a simple class balancing. The GAN augmentation approach is found to be subpar to classical methods for the considered dataset.
Paper Structure (14 sections, 5 equations, 16 figures, 3 tables)

This paper contains 14 sections, 5 equations, 16 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Materials and methods
  3. Dataset
  4. Image comparison metrics
  5. Classical metrics
  6. Fréchet Inception Distance
  7. Classical augmentation
  8. GAN-based augmentation
  9. Classification
  10. Saliency maps
  11. Results
  12. GAN-based image generation
  13. Classification results
  14. Conclusion

Figures (16)

  • Figure 1: A typical architecture of a Generative Adversarial Network.
  • Figure 2: Top row - example images from the dataset: a) normal, b) COVID-19 positive, c) lung opacity, d) viral pneumonia, e) lung segmentation mask. Bottom row - Original COVID-19 image (f) and the same image cropped and resized (g).
  • Figure 3: Examples of remaining annotations and marks on cropped and resized images.
  • Figure 4: Distribution of image similarity metrics: a) RMSE, b) SRE and c) SSIM. Four samples are compared to the COVID-19 sample: normal (yellow dotted), viral pneumonia (light blue vertical lines), lung opacity (red horizontal lines) and COVID-19 (blue diagonal lines).
  • Figure 5: Examples of GAN-generated images at different stages of the training process corresponding to decreasing FID values.
  • ...and 11 more figures