Integrating automatic and interactive methods for coronary artery segmentation: let the PACS workstation think ahead

Area of Science:

• Medical Imaging Analysis
• Cardiovascular Imaging Technology

Background:

• Coronary artery analysis is crucial for diagnosing heart disease.
• Existing methods for coronary artery segmentation and centerline extraction can be time-consuming and require significant user input.

Purpose of the Study:

• To introduce novel software for rapid coronary artery segmentation and precise centerline extraction.
• To enhance lesion visualization and quantitative measurement with minimal user interaction.

Main Methods:

• Optimization and integration of previously developed automatic and interactive coronary artery extraction techniques.
• Implementation of a user-friendly workflow with background processing for segmentation and centerline tracking.
• Development of an intuitive interactive analysis environment for immediate data review.

Main Results:

• Achieved an average overlap of 96.0% between software-generated centerlines and the reference standard.
• Demonstrated an average centerline distance of 0.38 mm compared to the reference standard.
• Automatic background processing completed in 1.4-2.5 minutes; interactive processing averaged 3 minutes.

Conclusions:

• The developed software offers higher efficiency compared to previous interactive methods.
• Preliminary results indicate reasonable accuracy for coronary artery analysis, comparable to manual extraction.
• The software streamlines the process of coronary artery segmentation and centerline extraction for clinical applications.