Article Lead Image

Tim Parkinson/Flickr (CC BY 2.0) | Remix by Jason Reed

This new smartphone app wants to help diagnose depression

Purple Robot is an Android app by a team from Northwestern University.


Dylan Love


Posted on Jul 16, 2015   Updated on May 28, 2021, 8:33 am CDT

The jury is still out on whether smartphones are making us happy or trying to kill us, but a new app claims it can identify when its users are unhappy, regardless of the cause.

A research team from Northwestern University developed an Android app called Purple Robot, which harnesses a smartphone’s sensor data to determine if the user is suffering from depression. Their findings, published in the Journal of Medical Internet Research, suggest that a depressed smartphone user’s behavior will ripple into and affect her smartphone’s own understanding of the world. A smartphone knows when it hasn’t left the house in three days. It knows when it hasn’t called anyone in weeks.

The team’s study looked at GPS and phone usage data of 28 participants over two weeks, and each participant took a PHQ-9 depression test, which is a self-test that tells you if you are at above-average risk of depression and does not formally diagnose anything. Based on smartphone data alone, Purple Robot successfully identified 87 percent of those that the PHQ-9 identified as being at risk.

Depression affects 6.7 percent of Americans over the age of 18, some 14.8 million individuals. Medical professionals often fail to identify its symptoms until it becomes a more serious issue. The promise of an app like Purple Robot is that it could passively and continuously monitor a user’s behavioral output, and thus tun into to any negative changes sooner and more accurately than a human physician who must see several patients a day.

The next step for Purple Robot is a larger study consisting of more participants over a longer time period. The aim will be to identify longer-term changes in behavior based on the smartphone’s accelerometer, GPS, and usage data; this early success is a positive indicator, but it’s an admittedly small study.

The app is not unlike Crowd++, a project from Rutgers University. They share many of the same potential pitfalls, too, including a lack of data whenever a user leaves his or her phone at home.

Still, the day could be near that our phones can tell us when we’re depressed (in case we weren’t already painfully aware). The next big question is how Northwestern’s researchers are going to cram a Baymax into the battery compartment so the phones can start to make us feel better, too.

H/T Al Jazeera | Photo via Tim Parkinson/Flickr (CC BY 2.0) | Remix by Jason Reed

Share this article
*First Published: Jul 16, 2015, 1:09 pm CDT