An Effective Deep Neural Network for Lung Lesions Segmentation From COVID-19 CT Images

IEEE Trans Industr Inform. 2021 Feb 12;17(9):6528-6538. doi: 10.1109/TII.2021.3059023. eCollection 2021 Sep.

Abstract

Automatic segmentation of lung lesions from COVID-19 computed tomography (CT) images can help to establish a quantitative model for diagnosis and treatment. For this reason, this article provides a new segmentation method to meet the needs of CT images processing under COVID-19 epidemic. The main steps are as follows: First, the proposed region of interest extraction implements patch mechanism strategy to satisfy the applicability of 3-D network and remove irrelevant background. Second, 3-D network is established to extract spatial features, where 3-D attention model promotes network to enhance target area. Then, to improve the convergence of network, a combination loss function is introduced to lead gradient optimization and training direction. Finally, data augmentation and conditional random field are applied to realize data resampling and binary segmentation. This method was assessed with some comparative experiment. By comparison, the proposed method reached the highest performance. Therefore, it has potential clinical applications.

Keywords: COVID-19; Conditional random field; data augmentation; deep network; lung lesions segmentation.

Grants and funding

This work was supported in part by National Natural Science Foundation of China under Grant 61701022, in part by National Key Research and Development Program (2017YFB1002804, 2017YFB1401203), in part by Beijing Natural Science Foundation (7182158), in part by Fundamental Research Funds for the Central Universities (FRF-DF-20-05), and in part by the Beijing Top Discipline for Artificial Intelligent Science and Engineering, University of Science and Technology Beijing.