Related Experiment Video
Updated: Jul 8, 2025
06:57
Author Spotlight: Advancing Cardiovascular Imaging - Introducing the Spatially Weighted Calcium Score for Early Disease Detection
Published on: September 22, 2023
1.0K
Spatial Attention-Guided Generative Adversarial Network for Synthesizing Contrast-enhanced Computed Tomography Images
Summary
This study introduces a novel spatial attention-guided generative adversarial network (SAG-GAN) to create contrast-enhanced CT images from non-contrast scans. This method aims to improve focal liver lesion diagnosis while reducing patient burden.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Radiology
Background:
- Contrast-enhanced computed tomography (CE-CT) provides vital information for diagnosing focal liver lesions (FLLs).
- CE-CT scans necessitate contrast agent injection, increasing patient burden and costs.
Purpose of the Study:
- To develop a method for synthesizing CE-CT images directly from non-contrast CT (NC-CT) images.
- To reduce the physical and economic burden on patients undergoing liver lesion diagnosis.
Main Methods:
- A spatial attention-guided generative adversarial network (SAG-GAN) was proposed.
- A spatial attention module was integrated into the generator to focus on relevant areas of the NC-CT image.
Main Results:
- The SAG-GAN successfully synthesized CE-CT images in both arterial and portal venous phases.
- Both qualitative and quantitative evaluations showed the SAG-GAN outperformed existing generative adversarial network-based methods.
Conclusions:
- The proposed SAG-GAN offers a promising approach for generating CE-CT images from NC-CT scans.
- This technique has the potential to improve FLL diagnosis efficiency and patient comfort.
