This paper proposes an automatic method for scapula bone segmentation from Magnetic Resonance (MR) images using deep learning. The purpose of this work is to incorporate anatomical priors into a conditional adversarial framework, given a limited amount of heterogeneous annotated images. Our approach encourages the segmentation model to follow the global anatomical properties of the underlying anatomy through a learnt non-linear shape representation while the adversarial contribution refines the model by promoting realistic delineations. These contributions are evaluated on a data set of 15 pediatric shoulder examinations, and compared to state-of-the-art architectures including UNet and recent derivatives. The significant improvements achieved bring new perspectives for the pre-operative management of musculo-skeletal diseases.