Fully automated 3D segmentation of MR-imaged calf muscle compartments: Neighborhood relationship enhanced fully convolutional network

• Published in: Computerized medical imaging and graphics Vol. 87; p. 101835
• Main Authors: Guo, Zhihui, Zhang, Honghai, Chen, Zhi, van der Plas, Ellen, Gutmann, Laurie, Thedens, Daniel, Nopoulos, Peggy, Sonka, Milan
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.01.2021
• Subjects: Calf muscle compartment segmentation; Edge constraint; Fully convolutional network; Humans; Image Processing, Computer-Assisted; Internal Medicine; Leg; Magnetic resonance image; Magnetic Resonance Imaging; Muscles; Neural Networks, Computer; Other; Edge constraint; 3D; Fully convolutional network; Calf muscle compartment segmentation; Magnetic resonance image
• ISSN: 0895-6111; 1879-0771; 1879-0771
• DOI: 10.1016/j.compmedimag.2020.101835

•We report a fully automated approach for 3D segmentation of five calf muscle compartments simultaneously.•By considering the similar textures shared by individual muscles, we are able to efficiently impose edge constrains in an end-to-end training manner.•Our method is robust in MR images from both healthy subjects and patients with DM1. Automated segmentation of individual calf muscle compartments from 3D magnetic resonance (MR) images is essential for developing quantitative biomarkers for muscular disease progression and its prediction. Achieving clinically acceptable results is a challenging task due to large variations in muscle shape and MR appearance. In this paper, we present a novel fully convolutional network (FCN) that utilizes contextual information in a large neighborhood and embeds edge-aware constraints for individual calf muscle compartment segmentations. An encoder–decoder architecture is used to systematically enlarge convolution receptive field and preserve information at all resolutions. Edge positions derived from the FCN output muscle probability maps are explicitly regularized using kernel-based edge detection in an end-to-end optimization framework. Our method was evaluated on 40 T1-weighted MR images of 10 healthy and 30 diseased subjects by fourfold cross-validation. Mean DICE coefficients of 88.00–91.29% and mean absolute surface positioning errors of 1.04–1.66 mm were achieved for the five 3D muscle compartments.