Related Experiment Video
Updated: Oct 16, 2025
08:05
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
14.4K
GACDN: generative adversarial feature completion and diagnosis network for COVID-19
Qi Zhu1,2, Haizhou Ye1, Liang Sun1
1College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, 211106, China.
BMC Medical Imaging
|October 22, 2021
Summary
This study introduces a generative adversarial network to create essential features for diagnosing COVID-19 from CT scans, improving accuracy and accessibility in resource-limited areas.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Radiology
Background:
- Coronavirus disease 2019 (COVID-19) is a global health crisis.
- Machine learning aids in diagnosing COVID-19 from chest CT scans.
- Current methods rely on time-consuming segmentation for feature extraction, limiting use in resource-scarce regions.
Purpose of the Study:
- To develop a novel method for COVID-19 diagnosis from CT images.
- To overcome limitations of manual segmentation in feature extraction.
- To improve diagnostic accuracy and accessibility, especially in areas with limited medical expertise.
Main Methods:
- A generative adversarial feature completion and diagnosis network (GACDN) was proposed.
- GACDN generates location-specific handcrafted features from radiomic features.
- Both original radiomic and generated handcrafted features are used for diagnosis.
Main Results:
- Generated features increased diagnostic accuracy by an average of 3.21% across four classifiers.
- The GACDN method demonstrated superior performance in COVID-19 vs. community-acquired pneumonia (CAP) classification.
- Experimental results validated the effectiveness on a COVID-19 dataset.
Conclusions:
- The GACDN method significantly enhances COVID-19 vs. CAP diagnostic accuracy with incomplete handcrafted features.
- The approach is suitable for regions with shortages of expert radiologists and high-performance computing.
- This facilitates wider application of AI in medical diagnosis.
