Tools & Strategies News

Researchers Use Artificial Intelligence to Predict Suicide Risk

A recent study described an artificial intelligence solution that can help determine suicide risks in women battling trauma-related disorders.

AI for risk predictions.

Source: Getty Images

By Mark Melchionna

- Researchers from Worcester Polytechnic Institute (WPI) and Harvard Medical School-affiliated McLean Hospital in Belmont conducted a study assessing the use of artificial intelligence (AI) to predict and gain a better understanding of suicidal self-injury.

According to the Centers for Disease Control and Prevention, the suicide rate for females in 2020 was 5.5 per 100,000. The overall suicide rate across both sexes increased by 30 percent between 2000 and 2020.

To combat growing suicide rates, the study published in the European Journal of Psychotraumatology aimed to determine whether AI could help predict women who are at risk of dying by suicide.

Researchers included 123 female patients at McLean Hospital in the study, 93 of whom had a history of child abuse, post-traumatic stress disorder (PTSD), and varying levels of dissociation. The study also included 30 controls.

The group of researchers developed an algorithm that was designed to predict suicide attempts among participants and identify subgroups of patients who were at the highest risk of entering a suicidal mindset.

They then used AI approaches to cluster data to expose any existing patterns. The patterns revealed a broad set of dissociative symptoms, such as the lack of connection between one’s sense of self and the environment. This was often due to trauma, the press release noted.

Following this step, researchers trained an algorithm to distinguish between patients with various dissociation levels and the 30 healthy controls. They found that the tool could zero in on specific dissociative symptoms, predicting previous suicide attempts with an accuracy of 83 percent.

Aside from AI being able to predict thoughts of self-harm and previous suicide attempts accurately, researchers also noted that the study emphasized the need for clinicians to assess patients for dissociative disorder symptoms.

“We’re trying to say that among these hundreds of symptoms and indicators, our results suggest these two or three symptoms may be helpful to focus in on,” said Dmitry Korkin, PhD, the Harold L. Jurist ’61 and Heather E. Jurist Dean’s Professor of Computer Science at WPI, in a press release.

Increasingly, AI is being used to power prediction models that can help improve healthcare delivery.

In October 2022, Penn State researchers said they plan to use a grant from the National Science Foundation to create AI-based machine-learning algorithms to analyze longitudinal data to predict health risks.

Researchers noted that the physiological parameters recorded when one seeks care at the hospital could be valuable in predicting health risks and observing how risks change with time. They plan to create machine-learning algorithms to analyze data and glean health risk insights.

Another effort in August 2022 involved the creation of an AI-based risk prediction model that used labor characteristics to determine potential child delivery outcomes.

Women in labor face clinical risks that are subjective and often unpredictable. To mitigate risks, researchers created the tool to leverage patient data collected at the start of labor, such as baseline characteristics, recent clinical assessments, and cumulative labor progress from admission.

After testing the algorithm on data from thousands of delivery episodes, researchers concluded that the AI risk prediction model was applicable and accurate.