Physics-based modeling of aortic wall motion from ECG-gated 4D computed tomography

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Computational Mechanics

Background:

• Electrocardiogram (ECG)-gated Computed Tomography (CT) provides 4D data for aortic wall motion analysis.
• Imaging artifacts like noise and motion blur limit the accuracy and reliability of 4D CT.
• Existing methods for motion correction may be inconsistent or difficult to implement across all image frames.

Purpose of the Study:

• To develop a physics-based filtering approach for constructing accurate dynamic models from 4D CT images of aortic wall motion.
• To address limitations posed by imaging artifacts in high-resolution 4D CT data.
• To enhance the reliability of aortic wall motion analysis using advanced computational techniques.

Main Methods:

• A physics-based filtering approach was developed to create a dynamic model from 4D CT images.
• A state filter was employed to correct simulated displacements from an elastic finite element model against observed image motion.
• An ensemble Kalman filter was used to refine model parameters, improving overall model quality.

Main Results:

• The proposed method successfully constructed dynamic models from 4D CT data.
• Performance was validated using synthetic data with known ground-truths.
• The approach was successfully applied to a real-world dataset, demonstrating its practical utility.

Conclusions:

• The developed physics-based filtering method effectively improves the accuracy and reliability of 4D CT for aortic wall motion analysis.
• This approach offers a robust solution for mitigating imaging artifacts in dynamic CT imaging.
• The technique holds potential for enhanced diagnostic capabilities in cardiovascular applications.