Point-Cloud Method for Automated 3D Coronary Tree Reconstruction From Multiple Non-Simultaneous Angiographic...

Area of Science:

• Medical Imaging
• Cardiovascular Imaging
• Computational Anatomy

Background:

• X-ray angiography is standard for coronary stenosis detection but suffers from observer variability in 3D geometry interpretation.
• Accurate 3D reconstruction of the coronary arterial (CA) tree is limited by motion and imaging artifacts.

Purpose of the Study:

• To develop and evaluate a novel 3D reconstruction method for the coronary arterial tree from 2D X-ray angiographic projections.
• To address limitations in accurate lesion severity determination caused by inter- and intra-observer variability.

Main Methods:

• Developed a new rigid and non-rigid motion correction technique for angiographic projections.
• Introduced a novel point-cloud based approach for 3D vessel centerline reconstruction via iterative error minimization.

Main Results:

• Achieved average reprojection errors of 0.092 ±0.055 mm for 3D centerline reconstruction.
• Demonstrated no statistically significant difference in luminal cross-sections compared to Optical Coherence Tomography (OCT) measurements.

Conclusions:

• The proposed method effectively reduces motion artifacts, enabling accurate 3D coronary arterial tree reconstruction.
• This approach offers a potential solution to improve the accuracy of coronary stenosis assessment and guide interventions.