Deep Transfer Learning for Interpretable Chest X-Ray Diagnosis

This work presents an application of different deep learning related paradigms to the diagnosis multiple chest pathologies. Within article, a well-known Convolutional Neural Network (DenseNet) is used and fine-tuned for X-Ray medical tasks. Different image augmentation methods are applied over training images improve performance resulting model as well incorporation explainability layer highlight zones picture supporting diagnosis. The finally deployed in web server, which can be upload get real-time analysis.