With depression rates around the world rising with each passing year, it has now become more relevant than ever to identity depression early and employ necessary methods to help individuals experiencing this state-of-mind. With this ideology in mind, researchers Andrew Reece from Harvard University and Chris Danforth from the University of Vermont have created a new algorithm that is supposed to help spot early signs of depression among people active on social networking sites such as Instagram.

Setting up the algorithm

Analyzing color patterns has always been a reliable way to understand a person’s psychological state, but can this be used to spot depression? Andrew and Chris employed the help of 500 workers from Amazon’s Mechanical Turk service in order to establish the answer to this question. The sample group was asked to answer a series of questionnaires that would help the researchers understand their tastes and preferences better, along with taking a standard clinical depression survey.

Out of the 500 people who participated in this study, 170 people agreed to share their Instagram photographs with the team. Of this diminished sample group, 70 people were found to be clinically depressed, and a total of 40,000 photographs were eventually put together. The participants were then asked to rate a large selection of pictures on a scale of 0 to 5 based on how interesting, happy or sad each picture appeared to them.

Color correction and correlation to depression

The results of the study were quite interesting and could play a major role in psychology in the future. The researchers noted that people who were depressed tend to put up images that featured a hue of blue, grey and generally darker colors. The most predominant filter used in this scenario was ‘Inkwell’. On the other hand, the average choice of filter by healthy individuals was ‘Valencia’, and they also ended up picking brighter and lighter colors in comparison.

People suffering from depression also ended up posting fewer photographs featuring faces, and their posts were also generally less popular online. When the algorithm was applied across random photographs, it had a 70 percent success rate in identifying individuals with depression.

This is a major step forward in fighting the manic wave of depression sweeping the world, and helping those in need before things escalate in a negative way.

Follow the page Tech
Don't miss our page on Facebook!