Related Experiment Video
Updated: Sep 30, 2025
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
Supervised and weakly supervised deep learning models for COVID-19 CT diagnosis: A systematic review.
Haseeb Hassan1, Zhaoyu Ren2, Chengmin Zhou2
1College of Big data and Internet, Shenzhen Technology University, Shenzhen, 518118, China; Guangdong Key Laboratory for Biomedical Measurements and Ultrasound Imaging, School of Biomedical Engineering, Shenzhen University Health Science Center, Shenzhen, China; College of Applied Sciences, Shenzhen University, Shenzhen, 518060, China.
Artificial intelligence (AI) and computer vision (CV) methods accelerate COVID-19 diagnosis using CT scans. Weakly supervised learning is more prevalent than supervised learning for COVID-19 CT image analysis.
Area of Science:
- Radiology
- Artificial Intelligence
- Computer Vision
Background:
- Artificial intelligence (AI) and computer vision (CV) offer reliable methods for extracting features from radiological images.
- These AI and CV methods can aid in early COVID-19 diagnosis, preceding traditional pathogenic tests, thereby saving critical time for disease management.
- This is particularly relevant given the global health challenges posed by the COVID-19 pandemic.
Purpose of the Study:
- To review and categorize deep learning-based COVID-19 computerized tomography (CT) imaging diagnosis research.
- To provide a baseline for future research by analyzing existing studies.
- To offer a unique multi-level arrangement of the collected literature, differentiating this review from previous works.
Main Methods:
- A systematic literature search was conducted across major databases (Google Scholar, IEEE Xplore, Web of Science, PubMed, Science Direct, Scopus).
- 71 relevant studies were identified and collected.
- The selected literature was classified into multi-level machine learning groups, including supervised and weakly supervised learning.
Main Results:
- The review found that weakly supervised learning approaches are extensively adopted for COVID-19 CT diagnosis compared to supervised learning.
- Weakly supervised methods, particularly conventional transfer learning, are effective for real-time clinical practices by leveraging sophisticated features without over-parameterizing models.
- Few-shot and self-supervised learning represent emerging trends aimed at addressing data scarcity and enhancing model efficacy.
Conclusions:
- Deep learning models, especially those employing AI, are crucial for effective COVID-19 disease management and control.
- Understanding deep learning approaches is essential for researchers and clinicians involved in ongoing COVID-19 CT diagnosis research.
- Weakly supervised learning offers a practical and efficient pathway for clinical application in COVID-19 diagnosis via CT imaging.
