Your smartphone senses your location and who you talk to when. But does can it detect when you’re feeling under the weather?
Anmol Madan explored this question in his thesis at MIT Media Lab. After completing a study that involved more than 320,000 hours of data from research participants’ mobile phones, he was able to model smartphone behaviors that predict the onset of common colds, depression, and influenza.
Now he and two other MIT alumni are using the research to launch a business. GINGER.io uses an Android app to collect SMS data, calling data and location data. When these behaviors change in a way that signals something could be wrong, it alerts the user.
Early stages of depression, for instance, often involve changes in how someone communicates. GINGER.io’s app, DailyData, picks up on those changes. In test deployments, the app was able to identify 60%-90% of the symptomatic days for mental health and common respiratory conditions. Theoretically, it will become better at doing so as more users opt to anonymously add their data to the pool for analysis.
“If you’re showing early signs of loneliness/depression, you might not report them to your doctor or family,” explains Madan. “The app currently detects these changes and sends alerts to you, but in the future, these alerts could be sent to a caretaker with your explicit permission.”
Users also have access to a dashboard that shows their baseline behavior and deviations from that baseline. It tries to predict when you might be symptomatic.
The startup used seed funding to launch with its first users in January, and it graduated from Boston TechStars earlier this month. Two medical providers are currently using the app with their patients.
Eventually, Madan hopes to pull in revenue from enterprises, providers and pharmaceutical companies that want to help their employees or patients stay healthy.
“We’re not a diagnosis,” he says. “We’re an early warning, self-support, self-serve tool.”