An innovative computer program, developed by researchers in the US, can diagnose early signs of depression in a person, simply by analyzing the photos he has posted on Instagram, Facebook and other social networks.
The program is based on the fact that when one starts having blackheads, this is reflected in one's photos.
Thus, it is possible - thanks to the appropriate algorithms of the program - to create a new type of early warning system for impending depression, which in the future could even be integrated into a mobile phone application.
The program so far, with 166 users and almost 44.000 photos, shows that it has a success rate of around 70% in the early diagnosis of depression, compared to 42% for physicians and general practitioners (not psychologists and psychiatrists).
The researchers, led by Chris Danforth and Andrew Reese of Vermont and Harvard Universities respectively, published the paper in the journal EPJ Data Science.
The study shows that the photos of users with onset of depression on social media, tend to have darker colors (a sign of negative mood), contain fewer faces and many lonely "selfies" (a sign of introversion), attract more comments from other users (not always positive) and make less use of the various filters.
When using a filter, someone like Inkwell on Instagram is more often used, which turns color photos into black and white.
Instead, mentally healthy users prefer filters like Valencia, which give photos a warmer and brighter tone.
In addition, people with depression are more likely to post photos than those who do not.
Source: Sigmalive