Using Deep Learning for categorizing radiographs of lungs. Implemented for Coding.Waterkant hackathon as a part of Waterkant Festival June 2020.

In the current critical situation of COVID-19 spread throughout the world appropriate image assessment can help to optimize treatment for patients admitted to hospitals. Currently imaging by x-ray radiography is standard, in uncertain cases with CT.

Patterns typical for COVID-19 commence with predominantly peripheral ground glass opacities visible on CT, followed by interstitial changes and consolidations that can become extensive at later disease stage, associated with a poor prognosis. Radiographs are less sensitive and specific compare to CT but still contain valuable information (W. Liang et al., JAMA 2020). Imaging may help improve patient stratification, e.g. predicting a poor outcome.

Used transfer-learning in Tensorflow. Experimented with VGG16, InceptionV3 and Xception model trained on ImageNet.

For training and validation of the networks a dataset of radiographs from a pre COVID-19 time Kaggle Chest X-ray Competition (Pneumonia) was used. This included 3883 cases with pneumonia and 1349 healthy controls.