Toward data‐efficient learning: A benchmark for COVID‐19 CT lung and infection segmentation

  • Jun Ma
    Department of Mathematics Nanjing University of Science and Technology Nanjing 210094 P. R. China
  • Yixin Wang
    Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences Beijing 100190 P. R. China
  • Xingle An
    China Electronics Cloud Brain (Tianjin) Technology CO., Ltd Tianjin 300309 P. R. China
  • Cheng Ge
    Institute of Bioinformatics and Medical Engineering Jiangsu University of Technology Changzhou 213001 P. R. China
  • Ziqi Yu
    Institute of Science and Technology for Brain‐inspired Intelligence Fudan University Shanghai 200433 P. R. China
  • Jianan Chen
    Department of Medical Biophysics University of Toronto Toronto ON M5G 1L7 Canada
  • Qiongjie Zhu
    Department of Radiology Nanjing Drum Tower Hospital the Affiliated Hospital of Nanjing University Medical School Nanjing 210008 P. R. China
  • Guoqiang Dong
    Department of Radiology Nanjing Drum Tower Hospital the Affiliated Hospital of Nanjing University Medical School Nanjing 210008 P. R. China
  • Jian He
    Department of Radiology Nanjing Drum Tower Hospital the Affiliated Hospital of Nanjing University Medical School Nanjing 210008 P. R. China
  • Zhiqiang He
    Lenovo Ltd. Beijing 100094 P. R. China
  • Tianjia Cao
    China Electronics Cloud Brain (Tianjin) Technology CO., Ltd Tianjin 300309 P. R. China
  • Yuntao Zhu
    Department of Mathematics Nanjing University Nanjing 210093 P. R. China
  • Ziwei Nie
    Department of Mathematics Nanjing University Nanjing 210093 P. R. China
  • Xiaoping Yang
    Department of Mathematics Nanjing University Nanjing 210093 P. R. China

抄録

<jats:sec><jats:title>Purpose</jats:title><jats:p>Accurate segmentation of lung and infection in COVID‐19 computed tomography (CT) scans plays an important role in the quantitative management of patients. Most of the existing studies are based on large and private annotated datasets that are impractical to obtain from a single institution, especially when radiologists are busy fighting the coronavirus disease. Furthermore, it is hard to compare current COVID‐19 CT segmentation methods as they are developed on different datasets, trained in different settings, and evaluated with different metrics.</jats:p></jats:sec><jats:sec><jats:title>Methods</jats:title><jats:p>To promote the development of data‐efficient deep learning methods, in this paper, we built three benchmarks for lung and infection segmentation based on 70 annotated COVID‐19 cases, which contain current active research areas, for example, few‐shot learning, domain generalization, and knowledge transfer. For a fair comparison among different segmentation methods, we also provide standard training, validation and testing splits, evaluation metrics and, the corresponding code.</jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p>Based on the state‐of‐the‐art network, we provide more than 40 pretrained baseline models, which not only serve as out‐of‐the‐box segmentation tools but also save computational time for researchers who are interested in COVID‐19 lung and infection segmentation. We achieve average dice similarity coefficient (DSC) scores of 97.3%, 97.7%, and 67.3% and average normalized surface dice (NSD) scores of 90.6%, 91.4%, and 70.0% for left lung, right lung, and infection, respectively.</jats:p></jats:sec><jats:sec><jats:title>Conclusions</jats:title><jats:p>To the best of our knowledge, this work presents the first data‐efficient learning benchmark for medical image segmentation, and the largest number of pretrained models up to now. All these resources are publicly available, and our work lays the foundation for promoting the development of deep learning methods for efficient COVID‐19 CT segmentation with limited data.</jats:p></jats:sec>

収録刊行物

被引用文献 (1)*注記

もっと見る

問題の指摘

ページトップへ