An Uncertainty-Driven GCN Refinement Strategy for Organ Segmentation

Roger David Soberanis Mukul1Orcid, Nassir Navab1,2, Shadi Albarqouni1,3
1: Technical University of Munich, 2: Johns Hopkins University, Baltimore, 3: Helmholtz AI, Helmholtz Center Munich
Publication date: 2020/12/11
https://doi.org/10.59275/j.melba.2020-8e2b
PDF · Code · Video · arXiv

Abstract

Organ segmentation in CT volumes is an important pre-processing step in many computer assisted intervention and diagnosis methods. In recent years, convolutional neural networks have dominated the state of the art in this task. However, since this problem presents a challenging environment due to high variability in the organ’s shape and similarity between tissues, the generation of false negative and false positive regions in the output segmentation is a common issue. Recent works have shown that the uncertainty analysis of the model can provide us with useful information about potential errors in the segmentation. In this context, we proposed a segmentation refinement method based on uncertainty analysis and graph convolutional networks. We employ the uncertainty levels of the convolutional network in a particular input volume to formulate a semi-supervised graph learning problem that is solved by training a graph convolutional network. To test our method we refine the initial output of a 2D U-Net. We validate our framework with the NIH pancreas dataset and the spleen dataset of the medical segmentation decathlon. We show that our method outperforms the state-of-the-art CRF refinement method by improving the dice score by 1% for the pancreas and 2% for spleen, with respect to the original U-Net’s prediction. Finally, we perform a sensitivity analysis on the parameters of our proposal and discuss the applicability to other CNN architectures, the results, and current limitations of the model for future work in this research direction. For reproducibility purposes, we make our code publicly available at https://github.com/rodsom22/gcn_refinement

Keywords

organ segmentation refinement · uncertainty quantification · graph convolutional networks · semi-supervised learning

Bibtex @article{melba:2020:001:soberanismukul, title = "An Uncertainty-Driven GCN Refinement Strategy for Organ Segmentation", author = "Soberanis Mukul, Roger David and Navab, Nassir and Albarqouni, Shadi", journal = "Machine Learning for Biomedical Imaging", volume = "1", issue = "MIDL 2020 special issue", year = "2020", pages = "1--27", issn = "2766-905X", doi = "https://doi.org/10.59275/j.melba.2020-8e2b", url = "https://melba-journal.org/2020:001" }
RISTY - JOUR AU - Soberanis Mukul, Roger David AU - Navab, Nassir AU - Albarqouni, Shadi PY - 2020 TI - An Uncertainty-Driven GCN Refinement Strategy for Organ Segmentation T2 - Machine Learning for Biomedical Imaging VL - 1 IS - MIDL 2020 special issue SP - 1 EP - 27 SN - 2766-905X DO - https://doi.org/10.59275/j.melba.2020-8e2b UR - https://melba-journal.org/2020:001 ER -

2020:001 cover