Oropharyngeal enlargement in obstructive lung disease: quantification and machine learning

Area of Science:

• Pulmonary Medicine
• Radiology
• Medical Imaging

Background:

• Limited understanding of upper airway remodeling in obstructive lung diseases (OLDs).
• Focus on lower airway remodeling in asthma and COPD contrasts with upper airway research gaps.
• Need for quantitative methods to assess upper airway changes in OLD.

Purpose of the Study:

• Investigate upper airway dimensions in OLD patients using quantitative computed tomography (QCT).
• Identify predictive parameters for OLD using machine learning techniques.
• Correlate upper airway metrics with lung function parameters.

Main Methods:

• Prospective cohort study including healthy controls, COPD patients, and asthma patients.
• Upper airway CT scans from oral cavity to subglottal region.
• Quantitative analysis using ITK-SNAP and QCT software, with machine learning for feature selection.

Main Results:

• Patients with OLD showed enlarged upper airway dimensions, including greater oral space air fraction and pharyngeal length.
• These metrics negatively correlated with forced expiratory volume in 1-second/forced vital capacity.
• Oral space air fraction and normalized pharyngeal length were key discriminators for OLD.

Conclusions:

• Quantitative CT and machine learning reveal oropharyngeal enlargement in OLD patients.
• Specific upper airway metrics can aid in predicting obstructive lung diseases.
• Oropharyngeal changes represent a significant aspect of upper airway remodeling in OLD.