Machine learning for the diagnosis of chronic obstructive pulmonary disease and photoplethysmography signal-based minimum diagnosis time detection

This study examined the feasibility of artificial intelligence-based diagnosis of COPD, using machine learning with the photoplethysmography signal and biomedical signal processing techniques. Included subjects (patients with COPD, n=8; healthy control subjects, n=6) were diagnosed as having or not having COPD by a specialist doctor using a Vitalograph Alpha spirometer.