Deep Learning for Chest X-ray

CheXNet Project - An ML project from Stanford ML group. Paper is available here. There are a few implementations of this work, but I like the one with Keras, which claims to achieve performance close to the original paper. In summary, the work uses DenseNet, a 121-layer CNN, to classify each reduce-sized input photo of size 224x224 (and probably 512x512 too).

The Stanford ML Group also accept volunteer projects about Healthcare and Education. Scroll to the end of this page.