Viewpoint planning for quantitative coronary angiography.

Area of Science:

• Medical Imaging
• Cardiovascular Interventions
• Computational Geometry

Background:

• Coronary angiography assesses myocardial blood supply using 2-D X-ray projections.
• Dimensionality reduction in X-ray imaging necessitates optimal viewpoints for accurate diagnosis.
• Current methods rely on physician's experience for selecting appropriate views.

Purpose of the Study:

• To introduce an algorithm for computing optimal viewpoints in coronary angiography.
• To enable reliable diagnosis by improving visualization of coronary arteries.
• To avoid the need for 3-D models in viewpoint planning.

Main Methods:

• Developed an algorithm for optimal viewpoint planning based on a single angiographic X-ray image.
• Computed subsequent viewpoints by rotating precisely around a vessel.
• Minimized vessel foreshortening in the computed views.

Main Results:

• The algorithm substantially reduces foreshortening compared to the initial view.
• Achieved complete elimination of foreshortening for specific rotations around isocentered vessels.
• Quantified worst-case boundaries for inaccuracies, enabling viewpoint design for desired requirements.
• Phantom study investigated the impact of input views on 3-D quantitative coronary angiography (QCA).

Conclusions:

• An algorithm for optimal viewpoint planning from single X-ray images was introduced.
• The computed viewpoints are guaranteed to reduce initial vessel foreshortening.
• The novel approach uses only fluoroscopy images, integrating seamlessly into clinical workflows.
• Implementation in QCA workflow enhances stability of vessel-shape reconstruction by standardizing viewpoints.