Artificial Intelligence Applications in the Management of Chronic Obstructive Pulmonary Disease: A Narrative Review Informed by a Systematic Search

Chronic Obstructive Pulmonary Disease (COPD) remains a leading cause of mortality, and early identification of phenotypes, exacerbation risk, and treatment response remains challenging. Artificial Intelligence (AI) has emerged as a promising tool for improving prediction, classification, and monitoring in COPD. This narrative review, informed by a systematic search, evaluated current AI applications in COPD management.

The review was registered in PROSPERO (CRD42024502765) and reported with reference to PRISMA 2020 principles. MEDLINE, Embase, and the Cochrane Central Register of Controlled Trials were searched from inception to May 2025. Eligible studies applied Artificial Intelligence, Machine Learning, or Deep Learning approaches to adult COPD populations for diagnostic, prognostic, monitoring, phenotyping, or disease management purposes. Data extraction was performed independently by two reviewers, and methodological quality was appraised using design-appropriate tools.

Twenty-two studies met the inclusion criteria. AI demonstrated high discriminative accuracy for COPD diagnosis, exacerbation prediction, and mortality risk (mean Area Under the Receiver Operating Characteristic Curve [AUROC] = 0.83), although methodological heterogeneity and limited external validation were common. Models integrating multimodal data generally outperformed single-domain approaches. Risk-of-bias assessment identified concerns related to small sample sizes, limited external validation, and incomplete reporting of model development and validation procedures.

Current evidence suggests that AI may support COPD diagnosis, prognostic prediction, and telemonitoring. However, methodological heterogeneity, limited external validation, and inconsistent reporting standards continue to limit clinical translation. Further high-quality, externally validated studies are needed to support implementation in respiratory care.