Mental illness has always been a serious problem within our community that has not been openly addressed until now. As more and more people opened up about the issues and called for action to help those in need, more and more ideas are rising to help the situation. In multiple startups and companies to battle mental illness, they involve artificial intelligence to help identify and help those suffering under depression, anxiety, and more.
X2AI’s Tess is a psychological AI that uses already existing messaging platforms such as WeChat and Facebook Messenger to conduct psychotherapy and provide psycho-education to a patient through their conversations. Although patients are talking to a machine, there are multiple positive qualities to this. Tess is available 24/7 and will not forget anything as a machine, nor will patients feel judged or fear social stigma. With 8 years of development, Tess’s emotion algorithm can accurately tell how a person is feeling and along with its conversation algorithm, it can easily understand negative expressions and alert for human intervention to take over the conversation.
Sometimes people don’t reach out about their issues or don’t even know it’s there in the first place. However, there are cues that can be picked up to recognize that mental illness is there. Using color analysis, metadata and algorithmic face detection, researchers from Harvard University and University of Vermont are utilizing machine learning and Instagram to detect signs of depression with a 70% accuracy rating. Some patterns they have noticed are that those with depression tend to post bluer, grayer, and darker colored photos. While they tend to avoid using Instagram filters, they seem to prefer to use the Inkwell filter to turn photos into black-and-white images.
Using transcripts and audio from psychiatric interviews, IBM research is using machine learning to help identify speech patterns typically found in patients with depression, schizophrenia, and more. What people say and write will be clear indicators of their mental health and provide early indicators of mental disease and help seek early treatment in the near future. IBM is currently creating an automated speech analysis application to run on mobile devises to provide a real-time overview of a patient’s mental health.