Cerebral Rolls Out Nationwide Patient Crisis Identification System

Digital mental health company Cerebral is using a machine learning algorithm to help pinpoint patients in crisis. And the recently unveiled initiative is “just the beginning” of Cerebral’s use of machine learning-enabled solutions, according to the company.

“Machine learning (ML) and artificial intelligence (AI) are critical tools in the advancement of mental health care, but these benefits are only possible at scale,” a team of Cerebral researchers explained in a company post. “Both technologies require many data points to test and validate hypotheses in order to prove that the systems are working effectively.”

Dubbed the “Crisis Message Detector 1 (CMD-1),” the newly touted tool is designed to identify messages from patients experiencing a mental health crisis, then refer those patients to a crisis response specialist.


Specifically, the tool was trained to spot signs of suicidal ideation, homocidal ideation, non-suicidal self-injury or domestic violence, according to the company post. If a patient’s messages to Cerebral have been flagged, a specialist will reach out directly and assess the patient’s risk level.

This mental health professional will then be able to call emergency contacts or local responders, if needed.

Cerebral specifies that it does not pitch this as an alternative to 911. The company claims the tool is accurate in properly identifying individuals in crisis.


“During a week-long pilot, CMD-1 screened over 60,000 EMR messages and flagged more than 500 potential crises,” the Cerebral researchers wrote. “The model successfully detected over 99% of all crisis messages and, as a result, crisis specialists were able to respond to patients in less than 9 minutes on average.”

Cerebral receives several thousand patient messages each day via its online chat system or mobile app. The company says that somebody from the patient’s care team reviews and addresses those messages, but that human-led process isn’t always as fast as it needs to be.

Photo credit: Cerebral

The company plans to expand its machine learning initiatives in the upcoming months and focus on issues such as response times for medication concerns, scheduling issues and general support requests.

“Because of Cerebral’s experience serving a quarter-million people (and counting), we are uniquely suited to develop and implement cutting edge ML/AI tools to supplement the expertise of our clinicians and help improve clinical outcomes,” the company’s post added.

CMD-1 is being rolled out nationally and will be available 24/7.

In addition to demonstrating the company’s interest in ML and AI, the patient-identification tool is also reflective of its previously discussed commitment to quality control moving forward.

Earlier this year, Cerebral came under fire for its prescribing practices of controlled substances. In June, news surfaced that the Department of Justice (DOJ) launched an investigation into the company based on a potential violation of the Controlled Substance Act.

In turn, the digital health company changed up its leadership team. Founder Kyle Robertson stepped down as CEO, and Chief Medical Officer Dr. David Mou took on the role.

“I will say that we have made mistakes,” Mou said at the American Telemedicine Association conference in May. “And I’ll also admit that we will continue to make mistakes and learn.”

Cerebral also announced layoffs for July 1. In June, Mou told BHB that the layoffs were reflective of the company’s priorities to keep behavioral health front and center while moving into value-based care.

As for the future, the company is looking to treat serious mental illness (SMI) and developing its value-based care proposition, Mou previously told BHB.

Companies featured in this article: