J. Pestian, M. Sorter, B. Connolly, K. B. Cohen, C. McCullumsmith, J. Gee, L.-P. Morency, S. Scherer, L. Rohlfs, A Machine Learning Approach to Identifying the Thought Markers of Suicidal Subjects: A Prospective Multicenter Trial, Journal on Suicide and Life-Threatening Behavior, 2016
Suicide is the deliberate self-inflicted act with the intent to end one’s life. By recent World Health Organization estimates, over 800,000 people die from suicide every year, with at least 20 times more attempted suicides. Despite the high cost to individuals, families, communities, and public health suicide remains a misunderstood and under-researched cause of death.
Predicting when someone will commit suicide is extremely difficult, but trained clinicians can identify the contributing factors to suicide risk using standardized clinical tools. Such tools can, however, be cumbersome and may not reliably translate into routine interactions between clinicians, caregivers, or educators. Automatic suicidal risk assessment based on behavioral analyses in short interviews could help to identify suicidal behavior efficiently and pervasively. To improve the success of suicidal risk classification and evaluation our work aims to understand and investigate the verbal and nonverbal behavior of suicidality.