Deep Learning-Based Airway Segmentation in Systemic Lupus Erythematosus Patients with Interstitial Lung Disease (SLE-ILD): A Comparative High-Resolution CT Analysis

A research team from Peking Union Medical College Hospital and Canon Medical Systems has developed an AI system that can automatically detect interstitial lung disease (ILD) in patients with systemic lupus erythematosus (SLE). The deep learning model, based on a customized U-Net architecture, analyzes high-resolution CT scans to segment airway structures at both lobar and segmental levels, providing precise volumetric measurements that reveal subtle structural changes invisible to the naked eye.

The study analyzed 106 SLE patients (27 with ILD, 79 without) and found statistically significant airway volume enlargement in specific lung regions. At the lobar level, the right upper lobe and left upper lobe showed significant differences between SLE-ILD and SLE-non-ILD patients. At the segmental level, specific segments including R1, R3, and L3 demonstrated the most marked changes, revealing a distinct topographic pattern of airway dilation in SLE-ILD patients.

This AI-powered approach represents a breakthrough in quantitative imaging biomarkers for autoimmune diseases. By providing objective, automated measurements of airway structures, the system enables earlier detection of ILD in lupus patients and allows for more precise monitoring of disease progression. The technology could transform how rheumatologists and radiologists manage SLE patients, moving from subjective visual assessment to data-driven clinical decision making.