AI-Driven Pulmonary Function Analysis for Respiratory Screening

The early detection and monitoring of chronic respiratory diseases such as Chronic Obstructive Pulmonary Disease (COPD) remains a major global healthcare challenge, particularly in primary care and low-resource settings. Pulmonary Function Testing (PFT) is considered the clinical gold standard for diagnosing respiratory conditions, yet traditional spirometry systems are difficult to deploy at scale due to their operational complexity, strict quality-control requirements, and the need for specialized interpretation. At the same time, recent advances in Artificial Intelligence (AI) and large-scale medical data platforms create new opportunities for intelligent, scalable, and automated pulmonary diagnostics capable of supporting clinicians in real-world settings.

This project proposes the development of an AI-powered pulmonary function analysis platform within the context of the China-Portugal AI Pulmonary Function Data Platform initiative, a collaborative effort involving Guangzhou Medical University, INESC-ID Portugal, and Macau University of Science and Technology. The proposed research will focus on the application of deep learning methods to respiratory signal analysis, including the automatic interpretation of spirometry curves and real-time quality control, anomaly detection. Inspired by recent advances in AI-driven healthcare systems, the project will explore machine learning models capable of reconstructing and predicting full pulmonary function patterns from short-duration expiratory signals, while correcting artifacts such as cough interruptions or premature termination of breathing maneuvers.