The goal of this research is to develop a wearable heart health monitoring system that is robust to motion artifacts. Currently, most wearable heart monitors detect the electrical activity of the heart, namely the electrocardiogram signal. However, the mechanical activities of the heart can also reveal valuable information about the health status of an individual. Commercial motion sensors, i.e. accelerometers and gyroscopes, can detect these activities by recording the vibrations induced by the heartbeat on the chest wall. However, the recorded vibrations often become contaminated by motion noise arising from sources other than the heart, such as the movement of the subjects. This research proposes a novel combination of hardware design and algorithm development to solve this problem as well as to provide more advanced heart health status metrics compared to current state-of-the-art wearable sensors. An array of motion sensors will be embedded in a wearable strap and placed around the chest wall. The algorithm takes advantage of the extra information provided by the sensor array to eliminate the motion noise in the recordings. Other types of sensors (such as electrical and optical sensors) will also be embedded in the system to augment the evaluation of heart wellness and provide detection of heart failure. Such a system will be invaluable for the home-based detection and management of cardiac diseases. The proposed research will be combined with various educational and outreach efforts. Specifically, the PI will engage high school students through the Liberty Science Center's "Partners in Science" program and continue to recruit undergraduate students, especially from female and minority groups, through the Summer Scholars Research Program at Stevens. The PI will also participate in providing data to Physiobank, which is the largest biophysical database online, available for the free use of the biomedical research community.
The goal of this research is to develop a wearable, multi-modal cardiovascular monitoring system with high robustness. There has been significant effort recently on the development of non-invasive wearable systems, in particular devices that monitor cardiac electrophysiology. However, in addition to the electrical aspects, a perspective on the mechanical activities of the heart and blood vessels also needs to be gained for a comprehensive evaluation of cardiovascular health. Progress in the area of wearable cardio-mechanical sensing is currently hindered by the challenge of overcoming motion artifacts. This research proposes a holistic hardware/software solution to this problem by implementing an array of inertial measurement units which are placed around the chest wall and record both the linear and rotational components of heart-induced chest vibrations. A model-based signal processing algorithm which takes advantage of the redundant information provided by the sensor array will then be applied to remove motion noise components from cardio-mechanical recordings in an embedded platform. The second thrust of this research is to augment the proposed cardio-mechanical sensing system with the more standard modalities of ECG and photoplethysmography (PPG). Sensor fusion algorithms will be used on the multi-modal signals to extract further cardiovascular features as well as increase the accuracy of the monitoring process. Finally, abnormality detection and classification algorithms will analyze the derived features to evaluate the wellness of the cardiovascular system and to detect heart failure. This research advances the state of cardio-mechanical sensing in terms of robustness and motion tolerance. Additionally, the proposed combined analysis of linear and rotational heartbeat-induced chest wall movements will provide a deeper insight into the theorem of cardio-mechanical sensing. Finally, the proposed noise cancellation approach provides an embedded framework for the extraction of weak biophysical signals from strong background noises and can be applied to other biophysical sensing modalities in similar application scenarios.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
