AI Aiding in the Diagnosis of Depression

An AI programme has been trained to search Facebook posts for what is known as“ red flags”, which can be indicative of sentences which emphasise on thoughts of loneliness, sadness, pain is being used to identify signs of depression. In fact, a US study has found that the AI managed to identify a case relating to depression more than three months earlier before health services found out.

Authors of the machine learning algorithm noted that during the testing phases of the algorithm as well as screening questionnaires which are used to identify early signs of depression. However, it has the upper-hand in being able to run without disruption in the background.

Following a recent surge in backlashes against social media giants like Facebook from prime ministers, popular figures and parents, many are becoming more aware and concerned about the detrimental effect on the wellbeing of children. This has led to Facebook being urged to implement age restrictions or at the very least, a time usage limit for younger people using the platform.

The researchers explained that the algorithm recognises these early warning signs such as feelings of loneliness or isolation. They elaborated on this by adding that the AI looks for keywords such as “alone”, “ugh” or “tears” as well as other minor factors like the timing and length of posts. Other clues that the AI takes into account include an increase in the use of first person pronouns like “I” and “me”. This method of screening could increase the likelihood of conditions being diagnosed, and treated, early, minimising the impact of depression on education, work and relationships.

AI being used to detect early signs of depression from social media posts

AI being used to detect early signs of depression from social media posts

The AI was put to the test by analysing 524,292 posts made by participants of a study on Facebook the team identified language markers indicative of early signs of depression. The algorithm was able to identify warning signs of depression in individuals from posts up to three months prior to their conditions being recorded in their medical records.

Although many say that social media poses a threat and has a severe effect on the mental health of children and young adults, technology like this could be used to take advantage of a growing problem. As this technology becomes more advances, we may be able to track signs of depression at a faster rate. 

Zacharia Sharif