Fig. 3.

A fully automatic workflow for assessment of fibrosis and quantification of scarred tissue. A convolutional neural network (CNN) makes segmentation possible without user intervention. Deep learning components are in yellow. Blue represents conventional image processing techniques and red illustrates assessment of fibrosis using LGE-CMR. Quality control assessment is performed after every step in the workflow. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)