A machine learning model that can identify people with post-traumatic stress disorder with 80% accuracy could become an screening tool to help health professionals. Inexpensive, the model candiagnose mental health disorders through telehealth platforms. "Sentiment analysis" involves categorizing data to see how many express positive thoughts and how many express negative thoughts. Using text gathered through 250 semi-structured interviews -- 87 with PTSD and188 people without PTSD -- the University of Alberta team identified individuals through scores indicating how often their speech featured mainly neutral or negative thoughts.

Read Full Story