An application of cascaded 3D fully convolutional networks for medical image segmentation

Our experiments illustrate the promise and robustness of current 3D FCN based semantic segmentation of medical images, achieving state-of-the-art results.1