Multi-stage transfer learning for lung segmentation using portable X-ray devices for patients with COVID-19
作者:
Highlights:
• Multiple transfer learning methodology able to work with a limited number of samples.
• Fully automatic proposal to segment the pulmonary region in low quality radiographs.
• Only fully automatic methodology designed to work with portable X-ray devices.
• Portable devices help to prevent contagion, and are the recommended during pandemic.
• Critical during the COVID-19 pandemic, as these devices return poorer quality images.
摘要
•Multiple transfer learning methodology able to work with a limited number of samples.•Fully automatic proposal to segment the pulmonary region in low quality radiographs.•Only fully automatic methodology designed to work with portable X-ray devices.•Portable devices help to prevent contagion, and are the recommended during pandemic.•Critical during the COVID-19 pandemic, as these devices return poorer quality images.
论文关键词:CAD system,Radiography,X-ray,Lung segmentation,COVID-19,Transfer learning
论文评审过程:Received 3 November 2020, Revised 2 January 2021, Accepted 30 January 2021, Available online 12 February 2021, Version of Record 24 February 2021.
论文官网地址:https://doi.org/10.1016/j.eswa.2021.114677