Microsoft looking to explore Wearable Blood Pressure Action via SmartBands

Microsoft Blood Pressure monitor wearable

Just when you thought that Apple and Samsung are going to lead the conversation around effective blood pressure monitoring via their existing smartwatch product lines, it appears that Microsoft is also exploring the whole personal health monitoring via wearable scene.

In what seems to be the first of probably many more signs to come, Microsoft is actively looking at some of the engineering around a wrist-based wearable that can monitor blood pressure.

Not surprising, given the interest of Satya Nadella around the future of healthcare and Artificial Intelligence.

Details of the Microsoft patent for blood pressure estimation

In a Patent approved today, 10,716,518, Microsoft’s technology licensing division breaks down its approach to Blood pressure estimation via a wearable computing device.

Hypertension, or high blood pressure, is a chronic condition that is generally asymptomatic but is a risk factor for a variety of medically significant outcomes, including stroke, heart failure, and coronary artery disease. 

It has a large impact on public health and healthcare spending, affecting between 16 and 37% of the world population. 

RELATED: 3 Best Blood Pressure Monitors to Use With Apple’s Health App

Microsoft’s method comprises training a machine learning model on a cohort data set. The cohort data set may include subject-specific contextual data, time-varying features, and blood pressure measurements for a plurality of subjects. 

In other words, Microsoft’s proposed Machine learning platform examines a person’s blood pressure data within the context of your medical history, co-relate with measurements of similar profiles, and produces an estimation of your Blood pressure.

For instance, knowing a patient’s heart rate, activity state (including recent activity levels), physical pose, time and type of last medication dose, or even static information from their health records (previous smoking history, recent significant weight loss or gain, pregnancy, etc.) can greatly augment the predictive power of a BP model.

Microsoft is focussing on a continuous Blood Pressure Monitoring solution (ABPM)

Patients often monitor blood pressure at home with an automated device, also with an inflatable cuff, which uses the oscillometric method of measurement, an approach with a similar principle that attempts to determine the changes in the pressure wave electronically (with some loss in accuracy). 

RELATED: WHOOP positioning as a clinical tool for Heart Failure management

These single-point measurements, however, do not necessarily predict whether a patient is truly hypertensive.

Blood pressure varies throughout the day, and a patient’s blood pressure is often higher at the doctor’s office at midday versus when relaxing at home, due both to the physical activity associated with travel to the clinic and the patient’s anxiety from meeting with the doctor (known as the “white coat effect”). 

An increasing body of evidence suggests that ambulatory BP monitoring (ABPM), continuous blood pressure monitoring provides a much more complete measure of blood pressure than either a single measurement at the clinician’s office or a single measurement with a home blood pressure device, and as such provides significant value in the diagnosis and treatment of hypertension.

How is Microsoft approaching Blood Pressure monitoring?

At the heart of the wearable is the embedded optical heart rate sensor (224). This sensor comprises an optical source configured to illuminate one or more blood vessels through a patient’s skin, and an optical sensor configured to measure reflected illumination from the blood vessels.

In this manner, measurements of the wearer’s heart rate, blood oxygen level, blood glucose level, or other biomarkers with optical properties may be generated. 

From a sensor perspective, this is very similar to some of the patents that we have seen in the past from other big tech players.

Unlike Samsung’s blood pressure monitor on its smartwatch, Microsoft’s process does not require periodic calibration by users. It is more aligned with Apple’s philosophy.

RELATED: Microsoft could be working on a brand new Microsoft Band

How is Microsoft’s proposed Blood Pressure estimation process different?

However, their approach takes into account other physiological conditions and uses them in a machine learning model in order to come up with a more accurate BP measurement.

Microsoft Blood pressure monitor wearable machine learning processMicrosoft Blood Pressure monitoring machine language

The wearable may detect a variety of conditions under which the patient’s blood pressure is likely to change

Based on detecting such conditions, the machine learning model may compare the blood pressure of the patient to the blood pressure that the patient had previously under similar conditions. 

RELATED: Is Apple’s approach to Blood Pressure monitoring via Apple Watch different than that of Samsung’s

For example, a large fraction of people exhibit a phenomenon known as “nighttime dipping,” in which the blood pressure drops significantly during the night. 

The wearable may detect a time of day and input the time of day into the machine learning model. The device can also be configured to detect when the patient is asleep, and when the patient is asleep detect a sleep stage. These detections of sleep and sleep stage may be used as inputs by the machine learning model in order to fine-tune BP readings. 

As with patents, this is still an abstract idea and we have no input whether Microsoft actively seeks to position itself in the Health Wearables segment anytime soon. It is however becoming apparent that more and more of the big tech players are actively evaluating the personal health space and remote monitoring opportunities.

As an Amazon Associate MyHealthyApple may earn commissions from qualifying purchases using our links

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.