A team of MU Electrical and Computer Engineering graduate students and faculty, alongside researchers from Nursing and Health Management and Informatics, recently reached a key breakthrough in the cuff-less measurement of relative systolic blood pressure.
The Mizzou team of researchers utilized non-contact hydraulic bed sensors to estimate relative systolic blood pressure — the top number in a typical blood pressure reading — by extracting features from the ballistocardiogram (BCG) signal. The correlation between their estimate and the ground truth blood pressure measurement using the typical cuff was 90 percent when compared to the BCG pulse strength and 83 percent when compared to the BCG pulse shape. This finding can be used to track changes in systolic blood pressure noninvasively as people sleep.
A ballistocardiogram signal represents the force exerted when the heart ejects blood into the great vessels of the body.
EECS graduate students Bo Yu Su and Moein Enayati and faculty Marjorie Skubic, Dominic Ho, Jim Keller and Giovanna Guidoboni were investigators on the project, “Monitoring the Relative Blood Pressure Using a Hydraulic Bed Sensor System.” They were joined by Laurel Despins and Marilyn Rantz from the Sinclair School of Nursing and Mihail Popescu of Health Management and Informatics.
The bed sensors, which are positioned under a mattress and are capable of measuring a subject’s pulse, captured the ballistocardiogram of 48 test subjects after they spent time riding on an exercise bike. The sensors tracked the change in the BCG signal over time, and the relative blood pressure results it indicated were compared to those gathered by the typical wearable cuff.
“Tracking of the systolic blood pressure is what geriatricians worry about for older adults,” Skubic said.
Researchers estimated the BCG pulse strength by computing the signal energy. And they extracted BCG signal shape by measuring the difference in the IJ and JK amplitudes in a BCG signal. The correlation between the BCG signal energy and cuff-based blood pressure measurement was 90 percent, while the correlation between the BCG shape and blood pressure was 83 percent.
With this method, it is feasible to use inexpensive, nonintrusive bed sensors to track changes in systolic blood pressure, making it a valuable method of monitoring the health of older patients.
“It very much builds on what we’re doing in being able to track health and identify health changes in older adults,” Skubic explained.