Your Instagram Posts May Offer Warning Signs of Depression
Technology and social media sometimes get a bad rap for addiction and anxiety to bullying for their negative impact on mental health, but the same tools can be used to help those in need.
Large-scale intervention programs such as the Crisis Text Line are expanding. The CTL not only enables trained counselors to support people in crisis situations 24 hours a day, 7 days a week (and free) via text messaging, but has also become one of the world’s largest health datasets and real-time crisis trending tracker. time. …
Smaller-scale research is also on the rise – Instagram is the newest service for capturing a picture of what is happening to users’ mental health.
According to a study published last week in EPJ Data Science, researchers at Harvard University and the University of Vermont wondered if markers of depression could be identified from photographs posted on the popular social network.
After using various computational techniques, including machine learning and image processing, to view nearly 44,000 photographs published by 166 study participants (identified as “healthy” or “depressed” based on previous clinical diagnosis), they answered yes.
The machines analyzed user activity, how many likes and comments the photo received, as well as the number of faces in the photo. Interestingly, depressed people posted more often, and although those posts received more comments, they received fewer likes. Their photos had faces, but there were fewer of them than those of the healthier participants.
The researchers also analyzed the hue, saturation, and value of the photo, and whether a filter was used (and what type of filter). Pictures posted by depressed people are usually per-pixel, bluer, darker, and grayer; and the filter was used less frequently than in healthy participants. When suppressed participants used filters, they preferred to convert color photographs to black and white.
Perhaps even more interesting is that the researchers were not only able to confirm depression in a person who had already received an official diagnosis, but the data were able to predict depression before the diagnosis was made.
In an interview with the New York Times, the researchers clearly stated that although the sample size was small, and from a very specially recruited group of active Instagrammers who were willing to share their past clinical diagnosis of depression, “the results are telling. to the promise of their methods. “