Title:The 2ST-UNet for Pneumothorax Segmentation in Chest X-Rays using ResNet34 as a Backbone for U-Net

Abstract:Pneumothorax, also called a collapsed lung, refers to the presence of the air in the pleural space between the lung and chest wall. It can be small (no need for treatment), or large and causes death if it is not identified and treated on time. It is easily seen and identified by experts using a chest X-ray. Although this method is mostly error-free, it is time-consuming and needs expert radiologists. Recently, Computer Vision has been providing great assistance in detecting and segmenting pneumothorax. In this paper, we propose a 2-Stage Training system (2ST-UNet) to segment images with pneumothorax. This system is built based on U-Net with Residual Networks (ResNet-34) backbone that is pre-trained on the ImageNet dataset. We start with training the network at a lower resolution before we load the trained model weights to retrain the network with a higher resolution. Moreover, we utilize different techniques including Stochastic Weight Averaging (SWA), data augmentation, and Test-Time Augmentation (TTA). We use the chest X-ray dataset that is provided by the 2019 SIIM-ACR Pneumothorax Segmentation Challenge, which contains 12,047 training images and 3,205 testing images. Our experiments show that 2-Stage Training leads to better and faster network convergence. Our method achieves 0.8356 mean Dice Similarity Coefficient (DSC) placing it among the top 9% of models with a rank of 124 out of 1,475.