Screening and staging of COPD with deep learning based on chest X-ray images and clinic

This multicenter retrospective study explores the potential of deep learning (DL) algorithms in diagnosing and assessing the severity of chronic obstructive pulmonary disease (COPD) using frontal chest X-ray (CXR) images and clinical data. The Ensemble model, combining CXR images and clinical parameters, demonstrated superior performance in distinguishing COPD compared to models using only clinical parameters or images. Additionally, it effectively predicted COPD severity with high accuracy. The study suggests DL algorithms as a promising tool for COPD diagnosis and staging, offering a sensitive and low-radiation approach for early detection and severity evaluation.