Loading web-font TeX/Main/Regular
Annotation-Efficient COVID-19 Pneumonia Lesion Segmentation Using Error-Aware Unified Semisupervised and Active Learning | IEEE Journals & Magazine | IEEE Xplore

Annotation-Efficient COVID-19 Pneumonia Lesion Segmentation Using Error-Aware Unified Semisupervised and Active Learning


Impact Statement:Chest CT has been proven a useful tool for diagnosis and assessment of progression of pneumonia findings associated with COVID-19. As the pandemic continues, fast and aut...Show More

Abstract:

As the coronavirusdisease 2019 (COVID-19) pandemic continues, fast and automatic COVID-19-related pneumonia lesion segmentation method is in an urgent need. The current s...Show More
Impact Statement:
Chest CT has been proven a useful tool for diagnosis and assessment of progression of pneumonia findings associated with COVID-19. As the pandemic continues, fast and automatic COVID-19 pneumonia lesion quantification and segmentation is in an urgent need. However, human expert annotation of such lesion required by current state-of-the-art methods is time-consuming and labor-intensive. In this article, we propose a method that alleviates the need for human annotation for high-quality pneumonia lesion segmentation. Compared with the conventional approach, the proposed method provides significant boost in performance (\sim10% in Dice score) given limited amounts of annotated data. The proposed method facilitates automatic diagnosis and quantification of COVID-19 pneumonia and has the potential to be applied on other diseases and diagnostic problems in the future.

Abstract:

As the coronavirusdisease 2019 (COVID-19) pandemic continues, fast and automatic COVID-19-related pneumonia lesion segmentation method is in an urgent need. The current state-of-the-art methods for segmentation generally require sufficient amounts of annotated data for training. However, human expert annotation of such lesion on chest computed tomography (CT) scans is time-consuming and labor-intensive due to its heterogeneous appearance, ambiguous boundary, and large number of slices in 3-D CT images. Therefore, the purpose of this study is to present a novel annotation-efficient learning method for COVID-19 pneumonia lesion segmentation on CT. To make the best use of limited human expert annotation resources, we propose an error-aware unified semisupervised and active learning method. A novel error estimation network is proposed to estimate a voxelwise segmentation loss map, which is used to guide learning from unlabeled data for semisupervised learning and choose the most informativ...
Published in: IEEE Transactions on Artificial Intelligence ( Volume: 4, Issue: 2, April 2023)
Page(s): 255 - 267
Date of Publication: 01 February 2022
Electronic ISSN: 2691-4581

Funding Agency:


References

References is not available for this document.