New Study Reveals AI Can Detect Subtle Facial Expressions Indicating Early Signs of Depression
A recent study has found that artificial intelligence (AI) can identify subtle indications in facial expressions that may signal early symptoms of depression, which can be difficult for the human eye to perceive. According to a research team from Japan's Waseda University, this technology could pave the way for simple and non-invasive screening tools for early detection of mental health issues in schools, universities, and workplaces, as reported in a study published on the scientific platform EurekAlert. The study, published in the journal Scientific Reports, used OpenFace 2.0, an automated facial expression analysis tool, to evaluate short videos of 64 students introducing themselves. Another group of 63 students was then asked to assess the degree of sympathy, gentleness, and naturalness of the individuals in these videos. The results showed that students who reported experiencing "subclinical" symptoms of depression (mild symptoms that do not warrant a clinical diagnosis but are considered a risk factor) appeared less expressive and less sympathetic to their peers. Interestingly, they were not perceived as more stressed or more awkward, but simply showed signs of a lack of positive emotion. However, the AI was able to detect very specific patterns in eye and mouth movements, such as a frown, a pursed lip, or a slight widening of the mouth. It found that these signs were strongly linked to depression levels, even if they are generally invisible to non-specialist observers. In this context, Dr. Eriko Sugimori, the study's lead researcher, stated: "The approach we developed, based on short videos and automated facial expression analysis, could become a simple and practical way to monitor mental health in educational and professional environments." Dr. Sugimori also explained that integrating this technology into digital health platforms or employee wellness programs could enable more proactive psychological monitoring and facilitate early intervention. The team emphasized that the ability to detect signs of depression before they become a full-blown clinical condition represents a significant opportunity to provide timely psychological support, reducing the risk of the disorder worsening. However, the researchers noted that the sample was limited to Japanese students, which calls for caution when generalizing the results. Indeed, different cultures influence the way emotions are expressed through facial expressions, and subtle signs related to depression may vary from one society to another. Despite this limitation, the researchers believe that the study's results clearly demonstrate the great potential of AI in the field of mental health. It offers a non-invasive, low-cost, and scalable tool for early detection of disorders.
