Lung Sound Analysis

This project develops an AI-assisted framework for digital lung-sound analysis to support respiratory assessment and physiotherapy monitoring. Auscultatory recordings will be collected using a digital stethoscope from six thoracic locations on the anterior and posterior chest, yielding 12 recording sites per participant. The signal-processing pipeline includes cardiac-sound suppression, denoising, respiratory-sound enhancement, and extraction of acoustic features. Machine-learning and deep-learning methods will be explored for disease classification and longitudinal assessment of changes following physiotherapy interventions. The project is conducted in collaboration with the Department of Physiotherapy, Faculty of Allied Health Sciences, and Department of Medicine, Faculty of Medicine, University of Peradeniya.