Non-invasive measurement of blood pressure with voice data

In North America, nearly half of adults experience chronically elevated blood pressure (BP), posing heightened risks of cardiovascular disease and kidney failure. Unfortunately, only half of those with elevated BP, or hypertension, are cognizant of their condition. The overarching goal of this project is to identify ways to monitor blood pressure with low cost, non-invasive and easy to use tools, thus providing an early warning mechanism to people developing hypertension. Gadgets that are currently available need frequent calibration and their accuracy often depends on such factors as skin tone and the handling of the device. To overcome these limitations, we aim to develop a measurement device that relies on voice data, gathered simply from standard phrases spoken into a mobile phone equipped with a custom app. We will use a combination of mathematical modelling and machine learning to robustly estimate blood pressure from voice segments. If successful, at-risk individuals would be offered a way to monitor their blood pressure that is as simple as making a phone call.

Faculty Supervisor:

Lennaert van Veen

Student:

Partner:

Klick Labs

Discipline:

Mathematics

Sector:

Health and Related Sciences & Technology

University:

University of Ontario Institute of Technology

Program: