A study shows the potential of large language models to detect signs of depression and suicide in patients
Large language models—artificial intelligence systems based on deep learning—could be useful in detecting mental health risks such as depression and suicide risk in narrative tests of patients undergoing psychiatric treatment. This is one of the conclusions of research published in JAMA Network Open, which also shows the potential of embeddings — a natural language processing technique that converts human language into mathematical vectors — to achieve the same end.