Related Experiment Video
Updated: Oct 10, 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
A coarse-refine segmentation network for COVID-19 CT images
Ziwang Huang1, Liang Li2, Xiang Zhang3
1School of Data and Computer Science Sun Yat-Sen University Guangzhou China.
IET Image Processing
|December 13, 2021
Summary
A new coarse-refine segmentation network accurately segments COVID-19 lung CT scans. This AI model improves boundary detection and multi-scale analysis for better infection assessment and treatment planning.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Pulmonology
Background:
- Accurate segmentation of COVID-19 infected regions in CT scans is crucial for effective patient treatment.
- Challenges include multi-scale infected areas and unclear boundaries due to low contrast in CT images.
Purpose of the Study:
- To develop an advanced segmentation network for precise COVID-19 CT image analysis.
- To address limitations of existing models in segmenting complex infected lung regions.
Main Methods:
- Proposed a novel coarse-refine segmentation network incorporating an atrous spatial pyramid pooling module.
- Utilized a hybrid loss function to enhance boundary definition and capture delicate structures.
- Implemented a coarse-refine architecture to improve segmentation accuracy.
Main Results:
- The proposed network demonstrated superior performance in segmenting COVID-19 CT images compared to established medical segmentation models.
- The model effectively handled multi-scale infected regions and improved boundary delineation.
- Achieved more accurate estimates of infection progression.
Conclusions:
- The developed coarse-refine segmentation network offers a significant advancement for COVID-19 diagnosis and treatment planning.
- Enhanced segmentation accuracy facilitates better clinical decision-making for COVID-19 patients.
- This AI-driven approach aids in providing more reasonable and effective treatment options.
