Accurate airway centerline extraction based on topological thinning using graph-theoretic analysis.

Area of Science:

• Medical Imaging
• Pulmonary Medicine
• Computational Anatomy

Background:

• Accurate quantitative analysis of the airway tree is crucial for diagnosing and treating pulmonary diseases using CT scans.
• Airway centerline extraction is essential for understanding airway structure, measuring parameters, and guiding visualization.
• Existing methods often produce inaccurate results, including extra branches and circular artifacts, hindering clinical applications.

Purpose of the Study:

• To develop an automatic and robust method for airway centerline extraction from CT images.
• To overcome limitations of traditional methods, such as false branches and circular artifacts.
• To improve the accuracy and reliability of airway analysis for clinical use.

Main Methods:

• A topological thinning method is employed to locate the initial centerline, preserving topological and geometrical properties.
• Graph-theoretic analysis is used to generate structural information of the airway tree.
• A distance weighting strategy and clinical anatomic knowledge are applied to remove artifacts like inaccurate circles and prune extra branches.

Main Results:

• The proposed method achieves over 96% branch identification accuracy.
• The technique demonstrates consistency across various clinical cases.
• The resulting airway centerlines are free from circular artifacts and extraneous branches, ensuring structural integrity.

Conclusions:

• The developed automatic centerline extraction method is robust and accurate for airway tree analysis.
• This approach significantly improves upon traditional methods by eliminating common artifacts.
• The findings support the use of this method for enhanced CT-based diagnosis and treatment planning for pulmonary diseases.