Peripheral bronchial identification on chest CT using unsupervised machine learning.

Area of Science:

• Medical Imaging
• Pulmonology
• Computer Vision

Background:

• Accurate identification of bronchial segments is crucial for diagnosing and managing lung diseases.
• Current methods for bronchial tree analysis can be labor-intensive and subjective.

Purpose of the Study:

• To develop and validate a robust, automated algorithm for detecting small- to medium-diameter bronchial segments.
• To improve the efficiency and consistency of bronchial tree analysis in computed tomography (CT) scans.

Main Methods:

• Segmentation of the peripheral pulmonary vascular tree to create cross-sectional images.
• Identification of bronchial center points using circular Hough transform and a novel variable ring filter.
• Application of agglomerative hierarchical clustering with 18 features to reduce candidate points and identify bronchial segments.

Main Results:

• Optimized algorithm parameters achieved >95% precision in identifying bronchial center points.
• The algorithm detected an average of 563 bronchial points per CT study with 96% mean precision across 21 datasets.
• Detection was robust across CT scans from two different vendors.

Conclusions:

• A robust algorithm for automatic detection of bronchial segments has been developed.
• The algorithm combines knowledge-based techniques with unsupervised machine learning.
• The method demonstrates consistent performance across different CT vendors and acquisition parameters.