IE 11 is not supported. For an optimal experience visit our site on another browser.

How three first-generation immigrants are using machine learning to improve mental health care

Spring Health's technology looks at the answers of a client assessment and recommends a treatment using a machine-learning algorithm.

April Koh remembers watching as her best friend tried seven different antidepressants while seeking treatment for severe depression starting in 2010.

After watching her friend struggle for two years and witnessing “how broken mental health care is today,” the entrepreneur knew she wanted to do something to help.

“Her psychiatrist said something like, ‘Let’s start with Lexapro’ the first time she saw him,” Koh recalled. “And we were both struck by how much it sounded like he was starting an experiment on her.”

Abhishek Chandra, April Koh, and Adam Chekroud, the three founders of Spring Health.
Abhishek Chandra, April Koh, and Adam Chekroud, the three founders of Spring Health.Chloe Tsang / NBC News

Abhishek Chandra, who met Koh while they were both studying at Yale, recounts a similar passion for mental health. While working at Google and at a medical startup, he was frustrated at the lack of technological innovation in healthcare.

“Healthcare has this interesting challenge where there’s so much tech that exists that people like Google are doing at least 10, 15 years ahead of mainstream healthcare. And that to me was infuriating and almost criminal, because the one place where you can most directly affect human life was not actually reaping the most immediate benefits of technology,” Chandra said.

I do believe that by making clinical assessment of patients universal and by increasing the effectiveness of treatments, one demonstrates to society that depression is a treatable illness, not a curse or a moral weakness

So when the pair came across Adam Chekroud’s research on depression in a scientific journal in January 2016, working together to launch a company focused on mental health seemed like the perfect fit.

In May 2016, the trio founded Spring Health in New York. The startup operates as an online mental health clinic for employers, using proprietary machine-learning technology to provide personalized treatment recommendations with hopes that they will be more effective.

Patients first complete a digital assessment that screens for mental health conditions. The assessment then generates personalized reports from their answers.

The reports and recommendations are based on technology developed using clinical research led by Chekroud, the company’s chief scientist and a current Ph.D. candidate in neuroscience at Yale.

In 2016, Chekroud published his findings from a two-year research study that used data from a clinical trial of depression to create a machine-learning algorithm that could look for patterns and predict successful treatments. His team then tested the algorithm in a separate clinical trial and found it still predicted effective treatment outcomes statistically reliably.

“We basically trained an algorithm to go through patients’ medical records and try and learn what kind of things indicate that someone is a good fit for medication or not,” Chekroud said. “Not only did our algorithm learn something about treatment outcomes in the first trial, but it seemed like it learned something true about depression outcomes in general.”

Spring’s data-driven approach to mental healthcare is based in the belief in “precision medicine,” an emerging medical model that encourages healthcare treatments tailored to patients’ unique attributes, Dr. John H. Krystal, chair of the Yale Department of Psychiatry and the chief of psychiatry at Yale-New Haven Hospital, said. He also serves as a member of Spring’s scientific advisory board.

“To my knowledge, Spring is the first practical application of the precision medicine approach in psychiatry,” Krystal wrote in an email. “This is a revolution for psychiatry because it has been a field that relied strongly on intuition rather than evidence. Increasingly, our valuable clinical insights and intuitions will be enriched by the conclusions derived from the analyses of clinical data.”

After receiving their personalized health reports, users have two treatment options: They can connect with one of Spring’s internal network providers, or they can seek outside care. If they choose to meet with a Spring provider, they do so virtually.

“This is really the next generation way of delivering mental healthcare. It turns out you don’t have to go to a therapist’s office in person to receive high-quality care. You can receive it via video,” Koh, Spring’s chief executive officer, said.

According to Koh, the approach gives Spring clients more flexibility in choosing a provider in addition to convenience. Because all therapy sessions are virtual, patients can choose to see a provider located anywhere in the country.

“Patients have really strong preferences around the race, gender, specialty and even geographical location of their provider,” Koh said. “And so for us, as we’re building our networks, one thing that’s incredibly important is diversity. We make sure we have a diversity of races and specialties.”

Koh also attributes the company’s commitment to diversity to their immigrant roots. All three company founders are first-generation immigrants.

“Our statuses as minorities in the States has really given us good perspective on what good mental healthcare looks like,” she said.

In addition to bringing greater accuracy and convenience to mental healthcare, Spring also hopes its data-driven approach will help reduce the stigma that surrounds public perceptions of mental health and encourage people to seek treatment.

Her psychiatrist said something like, ‘Let’s start with Lexapro’ the first time she saw him. And we were both struck by how much it sounded like he was starting an experiment on her.

Last month, fans across the globe mourned the loss of South Korean boy band member Kim Jong-hyun, whose apparent suicide has reinforced the necessity of addressing and improving mental healthcare. According to the World Health Organization, more than 300 million people around the world suffer from depression. And in United States alone, a 2015 study conducted by the National Institute of Mental Health found that an estimated that 16.1 million adults had at least one depressive episode over the course of the year.

“Most of current stigma arises from ignorance and misplaced fear. I do believe that by making clinical assessment of patients universal and by increasing the effectiveness of treatments, one demonstrates to society that depression is a treatable illness, not a curse or a moral weakness,” Krystal wrote.

“When we launch, what we find is that people start talking about mental health much more openly,” Koh said, adding that she's seen people who use the service discuss their experiences. "And so I can see how we are lifting stigma in a real way by having employers say, ‘We care about mental healthcare.’”

Follow NBC Asian America on Facebook, Twitter, Instagram and Tumblr.