Seismocardiographic Signal Variability and Pulmonary Phase Detection in Adults
Saved in:
| Title: | Seismocardiographic Signal Variability and Pulmonary Phase Detection in Adults |
|---|---|
| Authors: | Azad, Md Khurshidul |
| Committee Members: | Mansy, Hansen |
| Summary: | Cardiovascular disease is one of the leading causes of mortality in the world. Early detection and intervention can significantly improve disease management and patient quality of life. Current methods of evaluating cardiac function often involve history and physical examination (including stethoscope auscultation), electrocardiograms (ECG), echocardiogram imaging, computed tomography, and various blood tests. Seismocardiographic signals (SCG) are the chest surface vibrations resulting from cardiomechanical activity. SCG may be recorded using accelerometers and can be used for monitoring and predicting cardiac health. SCG's potential utility may be impeded by its spatial, postural, respiratory, and longitudinal variability. In this dissertation, the SCG variability sources are documented, resulting in changes in signal features are quantified, and optimum posture and sensor placement are discussed. Understanding SCG variability can help account for signal variability and more precise quantification of prominent SCG features that may be predictive of cardiac health. In addition, non-invasive monitoring respiration is a useful patient monitoring signal that can be performed via direct measurement of airflow utilizing a mouthpiece. In some instances, direct access to breathing airflow may be impractical or undesirable, especially in an ambulatory setting, and alternative approaches are needed. The respiratory phase can be extracted noninvasively from physiological signals such as ECG or SCG. The current study extracted respiration from several physiological signals in healthy adults and compared it with direct respiration airflow measurements. In addition, respiratory phases were extracted from SCG signals of HF patients, and results from traditional signal processing techniques and machine learning approaches were compared. The study resulted in a better understanding of the sources of SCG variability and alternative approaches to respiratory phase detection. These findings can lead to the development of improved non-invasive, low-cost methods for the management of cardiopulmonary conditions, timely intervention, and improved quality of life of patients. |
| URL: | https://stars.library.ucf.edu/etd2020/600 |
| Database: | OpenDissertations |
| Abstract: | Cardiovascular disease is one of the leading causes of mortality in the world. Early detection and intervention can significantly improve disease management and patient quality of life. Current methods of evaluating cardiac function often involve history and physical examination (including stethoscope auscultation), electrocardiograms (ECG), echocardiogram imaging, computed tomography, and various blood tests. Seismocardiographic signals (SCG) are the chest surface vibrations resulting from cardiomechanical activity. SCG may be recorded using accelerometers and can be used for monitoring and predicting cardiac health. SCG's potential utility may be impeded by its spatial, postural, respiratory, and longitudinal variability. In this dissertation, the SCG variability sources are documented, resulting in changes in signal features are quantified, and optimum posture and sensor placement are discussed. Understanding SCG variability can help account for signal variability and more precise quantification of prominent SCG features that may be predictive of cardiac health. In addition, non-invasive monitoring respiration is a useful patient monitoring signal that can be performed via direct measurement of airflow utilizing a mouthpiece. In some instances, direct access to breathing airflow may be impractical or undesirable, especially in an ambulatory setting, and alternative approaches are needed. The respiratory phase can be extracted noninvasively from physiological signals such as ECG or SCG. The current study extracted respiration from several physiological signals in healthy adults and compared it with direct respiration airflow measurements. In addition, respiratory phases were extracted from SCG signals of HF patients, and results from traditional signal processing techniques and machine learning approaches were compared. The study resulted in a better understanding of the sources of SCG variability and alternative approaches to respiratory phase detection. These findings can lead to the development of improved non-invasive, low-cost methods for the management of cardiopulmonary conditions, timely intervention, and improved quality of life of patients. |
|---|
