By Ben Hosking
How can we use data science and machine learning to prevent suicide? This question motivates Dr. Glen Coppersmith’s technology company, Qntfy. Dr. Coppersmith has had an interest in what he refers to as “the intersection of data and human behavior” ever since he received his BS in Computer Science from Northeastern University’s College of Computer and Information Science (CCIS), followed by a PhD in Psychology from Northeastern’s College of Science. Over 40,000 people die from suicide every year in the United States, and, when Dr. Coppersmith looked at the data space for mental health, he saw that there was not nearly enough solid data to make actionable inferences.
In January 2015, Dr. Coppersmith founded Qntfy, a company that works to empower human health and well-being. In mid-2016, his team of psychology-inclined computer scientists and clinicians started OurDataHelps, an initiative to get the systematic data needed for suicide prevention. OurDataHelps worked with thousands of volunteers who donated their individual social media data from platforms like Facebook, Instagram, and Twitter to Qntfy. The company’s researchers also looked at their data from other non-social platforms like fitness trackers and geolocation services, as well as personal histories via a questionnaire. These volunteer donors also self-reported any medical professional-diagnosed conditions such as bipolar disorder, schizophrenia, and post-traumatic stress disorder.
Qntfy looked at this vast collection of data using machine learning and pattern recognition to conclude that there are thousands of little signals that, when combined, can indicate a risk of suicide or self-harm. Qntfy uses a custom software built around deep learning and artificial intelligence. There was no single indicator, but, with their holistic approach, each data point provided a “brushstroke” to paint the bigger picture.
Aided by these data, Qntfy’s “biggest challenge is how to inform clinicians and integrate the data into healthcare outcomes.” According to Dr. Coppersmith, there is a pilot study with clinicians and real patients underway. He points out that “if we can identify people early enough, suicide prevention becomes a skill and resilience challenge” rather than a crisis challenge. “The more we can stay in preventative healthcare, the better.”
Dr. Coppersmith was recently featured on The Today Show to speak on Qntfy and OurDataHelps. When asked about how the morning news and talk show presented their data project on suicide prevention, he says, “they gave every bit of care to the subject as they could.” He wanted to create a positive way to share helpful information and says it’s “important to know how people with suicide experience would feel watching it.” The show also featured an individual with lived experience to help that purpose.
Dr. Coppersmith’s experience at Northeastern continues to shape his work and life. He remains friends with his advisor, Professor Richard Melloni, and praised the Northeastern co-op program. “Whenever a husky comes across my desk, I perk up,” he says. He links what he calls Northeastern students’ “pragmatism” to the co-op program’s practical real-world experience. His own co-op instilled in him the understanding that computer science is only a tool with which to solve problems. Even today, his ties to other alumni and the university offer him collegiality: “The Northeastern family remains. I immediately have a place to go back to at Northeastern.”