Related Experiment Video
Updated: Jun 4, 2025
07:10
A Sectioning, Coring, and Image Processing Guide for High-Throughput Cortical Bone Sample Procurement and Analysis for Synchrotron Micro-CT
Published on: June 12, 2020
5.0K
DDoCT: Morphology preserved dual-domain joint optimization for fast sparse-view low-dose CT imaging
Linxuan Li1, Zhijie Zhang1, Yongqing Li1
1School of Physics, Beihang University, Beijing, China.
Medical Image Analysis
|December 20, 2024
Summary
This study introduces DDoCT, a novel deep learning framework for low-dose computed tomography (CT) imaging. DDoCT effectively reduces radiation exposure while minimizing noise and artifacts for clearer diagnostic images.
Area of Science:
- Medical Imaging
- Radiology
- Artificial Intelligence
Background:
- Computed tomography (CT) is vital for medical diagnosis but raises public health concerns due to radiation dose exposure.
- Reducing radiation dose in CT involves lowering tube current or projection numbers, which can introduce noise and artifacts detrimental to image quality.
- Deep learning has shown promise in addressing challenges in low-dose CT (LDCT) imaging.
Purpose of the Study:
- To propose a dual-domain joint optimization framework (DDoCT) for reconstructing high-performance LDCT images.
- To mitigate noise from reduced tube current and streak artifacts from fewer projections in CT imaging.
- To enhance the applicability of LDCT in fast imaging environments.
Main Methods:
- Developed a dual-domain joint optimization framework (DDoCT).
- Applied DDoCT to noisy sparse-view projection data.
- Performed joint optimization in both the projection and image domains.
Main Results:
- DDoCT demonstrated significant progress in noise reduction.
- The framework effectively reduced streak artifacts common in sparse-view CT.
- Image contrast and clarity were substantially enhanced by DDoCT.
Conclusions:
- DDoCT offers a robust solution for high-performance LDCT imaging.
- The dual-domain optimization approach successfully addresses key limitations of dose reduction techniques.
- DDoCT shows great potential for practical fast LDCT applications.
