MIT app trials smartphone monitoring of hospital patients

Call it a startup made good., a health monitoring app created by the Massachusetts Institute of Technology, now has a clinical pilot underway to test its theory that smartphone use patterns can predict disease downturns.

A graduate of the TechStars mHealth incubator program, snagged $1.7 million in financing last fall to develop its product, which it now is testing with Cincinnati Children's Hospital Medical Center. The pilot project tests the technology on teens and young adults with inflammatory bowel disease (IBD), according to a report in Businessweek. Because the steroids that control the chronic condition can be harmful if taken for too long, clinicians hope to use the app to determine exactly when a patient's IBD symptoms subside, so they can discontinue the medications, the report says.

For example, the app might note when a patient becomes more mobile or leaves the house for the first time after an inflammatory episode begins. Researchers started with a handful of participants last fall, but are closing in on the 50 total they want for the full test, Michael Seid, lead researcher on the pilot, tells Businessweek. The app is "an effortless way to get a much finer-grained continuous measure of health status," he says.'s algorithms are built on a concept known as "machine learning," which sorts through smartphone calling, texting, messaging and other data to create a picture of the users' normal patterns. If the user's patterns change, such as a patient making no calls or texts for several days, the app first contacts the patient with a message asking about how he or she is feeling. If the anomaly is significant, however, the app can alert physicians or others in the user's support network. Creator Karan Singh describes it to Businessweek as a kind of human "check engine" light.

The app's developers also are working on a diabetes-specific module to see if they can detect depression among diabetic patients, which often causes patients to go off their medication. The prototype smartphone app will alert physicians when users start to withdraw or isolate themselves.

"People don't self-report that isolation," Singh tells Businessweek. "They don't necessarily say, 'Yes, I'm talking to less people this week,' or they don't necessarily realize that they've actually started to close themselves off."

Singh indicates the company may also be looking for ways to connect to biometric sensors that plug into smartphones as well, tracking more traditional remote patient monitoring data and adding to their algorithms to improve their predictive ability.

Other startups are testing the "passive monitoring" concept, too, to identify relapse triggers among drug addicts, and mobility problems among the elderly, Businessweek reports.

To learn more:
- read the Businessweek article
- check out iMedicalApps' interview with execs