Heart failure (HF) is one of the most major challenges faced by society today, claiming hundreds of thousands of American lives each year and costing more than 30 billion Medicare dollars annually. The ultimate goal of this research is to create a non-invasive and unobtrusive system for monitoring HF patients at home, automatically assessing their risk of experiencing an exacerbation, and providing feedback to caregivers and the patients themselves. This will enable proactive management of HF at home, with patients receiving tailored therapies that can adapt continuously to meet their changing needs. The hypothesis is that by (1) measuring a combination of hemodynamics, activity, and cardiovascular response to stressors (e.g. exercise) at home, and (2) combining these heterogeneous measures with modern data analytics, prediction of HF exacerbation at home can be achieved with a predictive window of greater than 7 days before impending hospitalization. To examine this hypothesis, the following four specific aims are proposed: (1) to collect longitudinal hemodynamic (ballistocardiogram, BCG) and activity data unobtrusively at home for the first time in a population of elderly HF patients; (2) to adapt existing algorithms fo predicting an impending HF exacerbation based on BCG, ECG, and activity time series data; (3) to develop wearable hardware based on BCG to be used at home for continuous hemodynamic and activity recording from elderly HF patients; and (4) to develop innovative sensing strategies and algorithms for improving the robustness of wearable BCG measurements. A previously demonstrated and verified weighing scale for center-of-mass (COM) BCG measurement will be scaled-up and deployed in the home of 200 total patients over the course of the project. Simultaneously, the hardware and analytics efforts will build on our prior data and existing prototypes. For the first 25 participants in the take-home BCG study, we will also conduct a direct physiologic study to quantify the underlying mechanisms contributing to both the scale-based (COM) and wearable BCG measurements, and to develop computational techniques for converting between the two domains (COM versus surface vibrations of the chest). These techniques, in combination with iterative and experimental efforts to improve the robustness of wearable BCG measurements, will then be applied to optimizing the wearable system for BCG and activity monitoring. This system will then be scaled-up and deployed in the home for the final 50 participants (of 200) in the take-home study. While we anticipate that the wearable will provide the best solution, the project risk is mitigated through the validation efforts with the existing scale-based system. Successful completion of this project could ultimately reduce HF related hospitalizations, and thus both improve quality of life for elderly Americans, and reduce overall healthcare costs.
Heart failure (HF) affects five million Americans and costs $30 billion dollars annually. The mainstay of HF treatment is monitoring of symptoms-a 'reactive' approach that is often too late to prevent a more serious outcome. We propose to evaluate a wearable, wireless, and internet-enabled technology that can monitor HF at home, predicting worsening of symptoms before they occur: a 'proactive' approach that will ultimately reduce both the incidence and cost of HF.
|Inan, Omer T; Baran Pouyan, Maziyar; Javaid, Abdul Q et al. (2018) Novel Wearable Seismocardiography and Machine Learning Algorithms Can Assess Clinical Status of Heart Failure Patients. Circ Heart Fail 11:e004313|
|Javaid, Abdul Q; Ashouri, Hazar; Dorier, Alexis et al. (2017) Quantifying and Reducing Motion Artifacts in Wearable Seismocardiogram Measurements During Walking to Assess Left Ventricular Health. IEEE Trans Biomed Eng 64:1277-1286|
|Ashouri, Hazar; Inan, Omer T (2017) Automatic Detection of Seismocardiogram Sensor Misplacement for Robust Pre-Ejection Period Estimation in Unsupervised Settings. IEEE Sens J 17:3805-3813|
|Ashouri, Hazar; Orlandic, Lara; Inan, Omer T (2016) Unobtrusive Estimation of Cardiac Contractility and Stroke Volume Changes Using Ballistocardiogram Measurements on a High Bandwidth Force Plate. Sensors (Basel) 16:|