Algorithms to prevent suicides

Listen to this article on the application
Listen to this article on the application

In the future, will we be able to predict suicide intentions as we anticipate showers? In the United States, algorithms are already trying to identify people at risk from digitized medical data. “A four-month information history is enough to establish a serious list of people who can act,” said Colin Walsh, a researcher at the American University Vanderbilt Medical Center in Nashville, Tennessee. The system, still experimental, has not been deployed in American hospitals. But it could ultimately serve as a screening tool for doctors.

Scientists are also looking at voice analysis to identify depressed people. They are also following the genetic trail: will they soon unravel the mystery of suicide by proving that it is written in our DNA? “This is only a research hypothesis”, tempers Jean-Pierre Soubrier, expert in suicidology for the World Health Organization. For several decades, this psychiatrist has accumulated books and studies on the subject, recently creating a documentation fund – one of the best endowed in the world – at the university library of Sainte-Anne hospital in Paris. “We know that there are families of suicides. The most famous is undoubtedly that of Hemingway. For the time being, we have not found ‘the’ chromosome that could explain everything”, adds the specialist. But research continues.

Analyze the smartphone of a suicide victim

“The evolution of technologies makes you want to do things in all areas, notes Christophe Debien, psychiatrist at the CHRU Lille. Currently, a study carried out in our establishment makes depressed patients listen to the voice they would have if they were were doing better, for therapeutic purposes. ” This example shows how artificial intelligence (AI) can revolutionize prevention. But this is a double-edged sword. “Imagine that we could retrieve and analyze the content of the smartphone of a suicide person. We would know his movements, his social interactions. From a scientific point of view, that would be a great wealth of information, that we could then use it to carry out prevention with very living individuals. But, on the other hand, constant surveillance would pose a real ethical problem, “underlines Christophe Debien.

Limited offer. 2 months for 1 € without commitment

For the time being, humans still largely take precedence over technology. “Even if we demonstrated the predictive potential of AI, it is not certain that the acceptability of professionals and patients would be good enough for the deployment of detection algorithms”, estimates Sofian Berrouiguet, psychiatrist at the CHU from Brest. But this does not prevent the implementation of innovative solutions.

Sleep analysis soon

Developed by the team of Prof. Philippe Courtet at the Montpellier University Hospital, the Emma application allows you to maintain a link with vulnerable people. When they are not doing well, they indicate it on their smartphone. The information is transmitted to a medical team which can then act, according to scenarios defined beforehand with the user. In the near future, sleep analysis could also be used to prevent suicides. “We know that the disorders of this state can be a precursor of the passage to the act. Like the cardiologists who monitor the blood pressure, the data collected during the phases of rest could therefore serve as markers”, even thinks Sofian Berrouiguet.

“From questionnaires and by looking closely at the risk factors (presence of a mental disorder or an addiction …), a doctor is able to tell if a person has a high risk of taking action within forty-eight hours, “says Christophe Debien. But what will happen in the two hours or the fifteen days which follow, when the patient will have returned home and that he will no longer be listening to the professionals? “The technology would make it possible to ‘fill’ these holes and to issue alerts for both patients and doctors,” explains Christophe Debien. A backup protocol defined in advance would then be implemented.

“The use of technologies to prevent suicides opens ethical debates of infinite complexity, adds Jean-Pierre Soubrier. But, beyond that, what will we do once we have detected and what are Our means There is no point in setting up these systems if we have not organized listening and follow-up, ”recalls the expert. Germany, like France, is in the process of thinking. The Land of Lower Saxony could soon test a video surveillance system based on artificial intelligence in order to reduce suicide attempts … in prison.





France, Paris, December 14, 2019, portrait of François Bazin.Francois Bazin


Each week, the Bookstore of the eco offers a title in the news.By François Roche


Frédéric Filloux.Frederic Filloux

Back to top button

Adblock Detected

Please consider supporting us by disabling your ad blocker