MIT Experts Develop Tool for Identifying Suicide Risk

MIT scientists have created a language-processing tool to assess suicide risk in mental health crises.

According to an initial, single report, scientists at MIT's McGovern Institute for Brain Research have developed an innovative tool to assess language during mental health crises and identify suicide risk. Created by Daniel Low and Satra Ghosh, the tool analyzes text conversations, looking for words linked to 49 identified suicide risk factors. They've tested it on a dataset provided by Crisis Text Line. About 16,000 de-identified conversations were grouped by risk levels: non-suicidal, suicidal ideation without imminent risk, and imminent risk, which includes plans or intentions to act within 48 hours. The tool assigns weights to factors, with mentions of lethal means and substance use often linked to the highest risk group.