Researchers testing AI-aided technique for heading off suicides

Researchers have developed a system combining AI and mHealth to catch signs of impending suicidal behavior in time for interventions by caregivers.

The team is recruiting psychiatric patients in France and Spain to participate in a study putting the system to the test in real-world settings.

The protocol for the study is running in BMC Psychiatry.

Advances in sensors technology and artificial intelligence methods (AI) “draw new perspectives for behavior ecological assessment,” explain Sofian Berrouiguet, MD, of Brest Medical University in France and colleagues. “Sleep disturbances and food intakes are not suicide risk factors usually screened in routine suicide risk monitoring,” yet “this clinical information could play a central role in suicide risk identification and prediction.”

Their system uses an app that collects data from smartphones’ native sensors to detect not only sleep trouble and appetite changes but also physical activity—or lack thereof—and other behavioral indicators that, taken together, can constitute red flags.    

Meanwhile a second app will perform “ecological momentary assessment.” This process involves assessing suicidal ideation via a self-questionnaire and a computer-assisted algorithm, the authors explain. Participants will answer structured questions twice daily plus one dynamic question asked at a random timepoint each day.

The system, which the researchers are calling Smartcrisis, could lead to a “paradigm shift from the traditional identification of risks factors to personalized prevention strategies tailored to characteristics for each patient,” they write.

To read the full study protocol, click here.