3D computed tomography airway geometry for predicting bronchoscopic accessibility in peripheral pulmonary nodules: a...

Area of Science:

• Pulmonary Medicine
• Medical Imaging
• Interventional Pulmonology

Background:

• Bronchoscopic diagnosis of peripheral pulmonary nodules (PPNs) using radial probe endobronchial ultrasound (rEBUS) is established.
• Predicting rEBUS accessibility for PPNs using computed tomography (CT)-derived airway geometry is underexplored.
• Accurate preprocedural planning is crucial for successful PPN diagnosis.

Purpose of the Study:

• To determine if quantitative CT airway geometric characteristics can predict rEBUS accessibility for PPNs.
• To assess if CT-derived airway analysis improves preprocedural planning for PPN diagnosis.
• To evaluate the utility of quantitative CT analysis in enhancing bronchoscopic accessibility predictions.

Main Methods:

• Prospective evaluation of rEBUS accessibility for 219 PPNs in 199 patients.
• Quantification of preprocedural airway geometry from CT scans.
• Development of predictive models using logistic-least absolute shrinkage and selection operator (LASSO) analyses, incorporating clinical and airway geometric variables.

Main Results:

• 83.1% of PPNs were easily accessible; 16.9% were difficult/inaccessible.
• Significant risk factors for limited access included acute bifurcation angles, sharply curved branches, and elliptical/narrower lumen shapes.
• A composite model (airway geometry + clinical variables) achieved the highest predictive performance (AUC = 0.84).

Conclusions:

• CT-based quantitative airway analysis significantly enhances the prediction of PPN accessibility to rEBUS.
• This approach supports more accurate procedural planning in clinical practice.
• Integrating quantitative CT airway data improves diagnostic yield for PPNs via rEBUS.