Image-to-Image Translation for Simplified MRI Muscle Segmentation

Deep neural networks recently showed high performance and gained popularity in the field of radiology. However, the fact that large amounts of labeled data are required for training these architectures inhibits practical application. We take advantage of an unpaired image-to-image translation approach in combination with a novel domain specific loss formulation to create an 'easier-to-segment' intermediate image representation without requiring any label data. The requirement here is that the task can be translated from a hard to a related but simplified task for which unlabeled data is available. In the experimental evaluation, we investigate the segmentation of pathological muscle tissue in T1 weighted magnetic resonance (MR) images of human thighs. The results clearly show improved performance in case of supervised segmentation techniques. Even more importantly, we obtain similar results with a basic completely unsupervised segmentation approach.