Many would argue that no night out is complete without a selfie. Usually multiple. In fact, usually an embarrassing amount that you rapidly erase from your Snapchat story the next morning before too many people see them. But what if our social media habits can be used by doctors to assess our mental health? Research shows that this development may not be too far in the future.
Researchers from Harvard University and the University of Vermont set out to analyse the Instagram posts of 166 individuals – equating to the investigation of 43,950 photos. These photos were analysed using a computer which looked at various variables, including brightness, colour, filter, number of faces and number of likes and comments. The mental health of the individuals was also assessed, and the results were very interesting.
“The photos were analysed using a computer which looked at various variables, including brightness, colour, filter, number of faces and number of likes and comments”
Those who were clinically depressed were much more likely to post photos which were bluer, greyer or darker. The most popular filter used in the depressed group was ‘inkwell’, which changes colour photos to black and white. Furthermore, there were more photos which contained faces but less faces per photo, meaning that selfies were more common amongst the depressed group. They were more likely to use Instagram more regularly and post more often when compared to the group who were not depressed. They also tended to have more comments on their posts, but less likes. This algorithm correctly predicted 70% of cases of depression, compared to an average of 42% by doctors after a one-on-one consultation. Amazing, right?
So what does this mean? We don’t fully know for sure just yet, although it’s certain that computers aren’t going to replace doctors in diagnosing depression anytime soon (despite the statistics suggesting otherwise). In reality, the population cannot really be separated into distinct categories like ‘depressed’ and ‘not depressed’ and so the usefulness of this programme is yet to be determined and is awaiting peer review. It also means that you shouldn’t panic if you’ve posted a few black and white selfies over the last few months. Although the uses of the algorithm have not fully been determined yet, it could aid in identifying those who need help and streamline the process between being diagnosed and receiving treatment.
Obviously, there are privacy issues when it comes to this type of technology; although much of social media is public, not everyone would be happy with their doctor using an algorithm on their posts to determine their mental health. It’s safe to assume that there is still a long way to go before this kind of technology is fully integrated into the world of health – but with today’s machinery and tech habits advancing so quickly, I say that it’s about time medicine caught up.