Ballistocardiogram-Based Approach to Ubiquitous Blood Pressure Monitoring

Abstract

Hypertension is a large contributor to cardiovascular deaths in the United States and around the world. Despite this fact, lack of ubiquitous blood pressure monitoring remains a large gap in modern-day health care. This thesis shall present a ballistocardiogram-based approach to ubiquitous blood pressure monitoring. Two approaches are presented: a signal-based approach and a model-based approach. The signal-based approach validates the superiority of ballistocardiogram-based pulse transit time as a reliable marker for BP over current electrocardiogram-based pulse arrival time methods. Concepts from the signal-based approach are then combined with a mechanism model of the ballistocardiogram and arterial hemodynamic modeling methods to produce a model-based approach. Two model-based variants are presented and the results of each are discussed. Despite certain limitations of the model-based approach, both variants are shown to still be more reliable than electrocardiogram-based pulse arrival time for blood pressure prediction. This work provides the foundation for a truly ubiquitous blood pressure monitoring method and paves the way for early detection and continuous monitoring of cardiovascular health issues.