GANs 'N Lungs: improving pneumonia prediction

Tatiana Malygina, Elena Ericheva, Ivan Drokin

Keywords: X-ray medical image classification, data augmentation, CycleGAN

TL;DR: We propose CycleGAN-based augmentation method to improve model performance in binary classification tasks in biomedical field

Abstract: We propose a novel method to improve deep learning model performance on highly-imbalanced tasks. The proposed method is based on CycleGAN to achieve balanced dataset. We show that data augmentation with GAN helps to improve accuracy of pneumonia binary classification task even if the generative network was trained on the same training dataset.

Code Of Conduct: I have read and accept the code of conduct.

Remove If Rejected: Remove submission from public view if paper is rejected.