Early detection of pathological behaviors by studying heart sounds

Abstract

This thesis investigates advanced signal processing techniques for the early detection and management of cardiac pathologies, focusing on the analysis of phonocardiogram (PCG) signals and intracardiac pressure estimation. It emphasizes the significance of accurate cardiovascular health assessment as a critical factor in improving patient outcomes. The research employs artificial intelligence (AI) methodologies, leveraging machine learning algorithms to enhance the accuracy of pathological identification in PCG signals. Additionally, it explores non-invasive methods for estimating cardiac pressures, highlighting their crucial role in diagnosing conditions such as heart failure, valvular disorders, and pulmonary hypertension. The findings demonstrate that integrating AI with traditional cardiac monitoring can significantly improve diagnostic precision, facilitating timely clinical interventions and paving the way for more effective cardiovascular disease management.