Combining Shape Priors with Conditional Adversarial Networks for Improved Scapula Segmentation in MR images

TL;DR

This paper introduces a novel deep learning approach combining shape priors with adversarial networks to improve scapula segmentation in MR images, especially with limited annotated data, enhancing clinical pre-operative planning.

Contribution

It integrates anatomical shape priors into a conditional adversarial network framework, improving segmentation accuracy over existing methods with limited data.

Findings

Significant improvement over UNet and derivatives

Effective incorporation of anatomical priors

Potential for better pre-operative planning

Abstract

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 dataset 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.