Catching the silent killer
High blood pressure — often called a “silent killer” because it rarely produces symptoms — can be accurately diagnosed by examining two commonly recorded biosignals, research from the University of British Columbia and the Guilin University of Electronic Technology in China suggests.
As blood flows through the body, it creates signals that include a wealth of information about the state of the heart and arteries. The research team identified several “signatures” within two types of these signals that collectively indicated whether or not a patient had high, or higher than normal, blood pressure with up to 95 per cent accuracy — even when the signals were not of optimal quality.
“Taken together, these signatures could form the basis for a quick, painless, and reliable monitoring method that enables people to detect and seek treatment for hypertension earlier than they otherwise might,” says Mohamed Elgendi, an adjunct professor of electrical and computer engineering at UBC and a postdoctoral fellow at British Columbia Women’s Hospital, who led the study.
Cardiovascular disease kills one out of every three people worldwide, and its primary cause is high blood pressure. But due in part to limitations with current monitoring methods, many people do not know they are hypertensive until it’s too late. Intra-arterial blood pressure monitoring — the current gold standard — is an invasive procedure that requires a tube to be inserted into an artery, while blood pressure cuffs not only are cumbersome to use, but also have been shown to produce inaccurate results.
Fortunately, there is another, noninvasive way to monitor blood pressure: recording and deciphering biosignals such as electrocardiograms (ECGs) and photoplesythmograms (PPGs), which change as blood pressure changes. When the team analyzed 121 sets of ECG, PPG, and intra-arterial blood pressure readings — all gathered from MIT’s Medical Information Mart for Intensive Care (MIMIC) database — they found that 10 signatures within PPGs revealed whether or not a patient had high blood pressure with almost 95 per cent accuracy when combined with a single signature within ECGs.
The 11 signatures also showed whether or not a patient was prehypertensive, or had higher than normal blood pressure, with 84 per cent accuracy. Greater accuracy may be achieved by applying the method to biosignal recordings that are at a higher frequency and level of synchronization than the MIMIC database currently provides.
“PPGs and ECGs can now be recorded using smartphones, which are accessible even in under-resourced communities,” notes Elgendi. “My hope is that by incorporating our findings into these devices, they can be used to diagnose hypertension rapidly and accurately and ultimately reduce the number of hypertension-related illnesses and deaths around the world.”