Monitoring wellness with passive sensors

Monitoring wellness with passive sensors

Note: This post originally appeared on Martin Sumner-Smith‘s Digital for Health blog. It has been reposted here with his permission.

While ‘standard smartphone features’ such as voice recording, motion-sensing and personal geolocation have found surprisingly powerful application to mobile health needs, specialized sensors are going to be required to address most medical applications. It simply isn’t possible to measure all key biological parameters without them.

To get an idea of what might be measured, consider current medical tests. Some such tests are performed by individuals already – the measurement of blood sugar levels by diabetics is an obvious example. Similarly, blood pressure, heart rate, body weight and height (to derive body mass index – BMI) are commonly measured. But in each case, these measurements require that a person manually perform them. As a result, they may not be performed when they should be, or at the optimal frequency.

Passive sensors, likely linked to an application on a smartphone, can address this issue. They enable the automation of measurements while not requiring any lifestyle changes by the user.

The following image links to an interesting infographic that originally appeared in the paper Making Sense of Sensors: How New Technologies Can Change Patient Care written by Jane Sarasohn-Kahn. It gives some passive sensor examples, both used on a person’s body and installed in their environment:

Kahn notes that passive sensors will be used first by people who simply want to stay well, and that adoption depends on their passive nature – monitoring occurs automatically without action by a user.

But there is another significant benefit – each measurement is captured and stored. In this way, developing trends can be assessed over time and sporadic, acute events are more likely to be captured. As I have noted in previous posts, each parameter being measurements can be correlated with other parameters such as where the person was (i.e., what were the environmental conditions), what else they were doing (e.g., exercising, driving, eating, etc.), the time of day, etc. The correlation of multiple parameters can elucidate triggering events, as well as beneficial and detrimental health factors. More sophisticated and beneficial health monitoring is possible with a larger number of frequently measured parameters.

It is obviously better that a person monitor their own blood pressure on a regular basis than it is for them to go to a doctor’s office to have it monitored, provided they can do so accurately. Automation can assure accuracy.

But blood pressure is one of the few medical tests currently performed by individuals themselves (others were noted in the introduction). Most medical tests  are expensive, performed very occasionally in batches in a laboratory setting, and only when requested by a physician. However, the benefits of more frequent measurements recorded over time and correlated with other parameters apply just as much to these tests, but cannot be realized at present.

When effective, passive sensors become cheaply available for a wider range of medical tests, we are likely to see considerable disruption in health delivery, not the least of which will be the disintermediation of medical testing.