Cardiac motion correction based on partial angle reconstructed images in x-ray CT

Area of Science:

• Medical Imaging
• Computed Tomography
• Cardiovascular Imaging

Background:

• Cardiac CT imaging faces challenges due to significant heart motion, even with high-speed scanners.
• Existing motion compensation algorithms often require extensive projection data or reconstructed images for accurate motion estimation.
• Minimizing scan range is crucial for reducing radiation dose in cardiac CT.

Purpose of the Study:

• To develop a novel cardiac motion estimation and compensation algorithm using a sinogram with a rotation angle less than 360°.
• To improve the quality of cardiac CT images by accurately estimating and compensating for heart motion.
• To reduce radiation dose by minimizing the scan range required for motion estimation.

Main Methods:

• Reconstruction of two 3D partial angle reconstructed (PAR) images from a limited angular range (180° + α + β).
• Estimation of a whole-heart motion model via nonrigid registration between the PAR images after artifact preprocessing.
• Motion-compensated image reconstruction at a target phase using the estimated motion model.

Main Results:

• Motion-compensated 3D images demonstrated significantly reduced motion artifacts in coronary arteries.
• Restoration of object boundaries affected by motion was observed.
• Improved visibility of coronary arteries and overall enhanced visual quality in both digital and physical phantom studies.

Conclusions:

• A novel PAR image-based cardiac motion estimation and compensation algorithm has been developed.
• The algorithm successfully operates with an angular scan range of less than 360°.
• The proposed method shows excellent performance in reducing cardiac motion artifacts, validated by phantom datasets.