Dartmouth researchers have developed the first smartphone application that uses AI and facial-image processing software to predict depression before users become consciously aware of symptoms.
The researchers found that the app, MoodCapture, could predict depression’s onset with 75% accuracy.
Here’s how it works.
Every single time you open your phone with facial recognition, the app measures your gaze, the position of your head, muscle rigidity, and other specific facial expressions and environmental cues.
It then pairs that with deep-AI learning over time about your normal gaze, rigidity etc., to make the prediction.
Of course, it’s much more complex than that, but you get the gist.
The goal, then, is to alert users that they may have depression lurking at the door and — armed with that knowledge — we can then engage in whatever therapeutic measures we’ve found work best. Reach out to a friend, therapist, go for a run, any of those kinds of things.
I’ll admit, sometimes things like this feel dystopian and creepy – the camera, particularly. The way the app can predict you.
But at the same time, maybe that’s just me getting older and some people (including myself) could probably really benefit from it.
If you notice, there have been quite a few AI-related advances in this field (including that remarkable study showing AI could predict the effectiveness of Zoloft after just a week).
The medicinal benefits of AI are only just beginning.
[Photo: Pexels, free stock photography]
