Every month we write about new technologies developed to monitor blood pressure. High blood pressure continues to plague Americans in massive proportions, affecting two-thirds of the U.S. adult population if you count both hypertension and prehypertension, says the CDC. This month already we’ve written about a new smartphone sensor and a wearable ultrasound patch, both presented as the best bet yet for accurate and dependable blood pressure monitoring. But what if all those development teams are working on the wrong side of the elephant?
Researchers at the University of California San Diego’s Jacobs School of Engineering are using tech to combat hypertension in a novel approach that is unrelated to measurement tools. We know lifestyle choices affect blood pressure for better or for worse. What if the first step in any hypertension treatment plan were to identify and focus on the one lifestyle factor that makes the greatest difference in blood pressure on a personalized basis?
The UCSD engineers demonstrated machine learning’s potential to predict an individual’s blood pressure and provide personalized treatment recommendations. Citing the state of overwhelm that descends on patients when doctors advise them to make a slew of significant lifestyle changes, the team took another course. The engineers collected sleep, exercise, and blood pressure data from eight patients, which is an admittedly small sample. The patients used FitBit Charge HR and Omron Evolv wireless blood pressure monitors to measure and report the data. The UCSD team developed an algorithm that predicts an individual user’s blood pressure and which behavior affects it the most. For one person the most significant factor might be sedentary minutes per day, while a different individual’s blood pressure changes most based on the number of hours of sleep.
The UCSD team published its work so far in a report that was deemed the Best Paper at IEEE Healthcom 2018. Future work involves testing with larger sample patient populations, generating daily one-day-ahead predictions, and a larger study of the long-term effect of lifestyle behaviors on blood pressure.
