Microsoft and SilverCloud collaboration to leverage AI for mental health

The treatment of mental health conditions appears to have received a boost with a recently announced research collaboration between digital mental health company SilverCloud Health and Microsoft Research. The partnership was designed to further step up the former’s online offering with artificial intelligence.

A little background: During the past 18 months, the two have worked in tandem on research that marries Microsoft’s machine learning and AI technologies with SilverCloud, which specializes in the digital delivery of evidence-based mental healthcare to improve outcomes.

Ken Cahill, CEO of Boston based SilverCloud, said the technology enables “very tailored support” to each patient; meaning “more responsive and reactive care.” He called that process a “big departure” from existing digital delivery that’s generic or a one-size-fits-all approach and doesn’t accommodate for factors such as behavior, engagement, and effectiveness.

SilverCloud’s digital mental health platform is globally deployed in routine clinical care providing coverage to 65 million people, Cahill noted. Since 2012, more than one million hours of therapy have been delivered.

“Our (model) encourages patients to be more active (in the care) instead of a backseat spectator. We’re all individuals with unique challenges and underlying issues,” noted Cahill. He added that the goal of the AI collaboration with Microsoft is to provide personalized mental health care which will hopfeully improve mental healthcare outcomes globally.

Artificial intelligence necessarily evokes the feeling of machines doing the work that humans did, so Cahill emphasized that SilverCloud views itself as “an extender – not a clinician replacer.”

In other words, “we’re working with Microsoft on ensuring what we deliver is responsive, in context and appropriate for that end user.” He said that if you contemplate a user’s journey today, they’re accessing what’s often “generic care, that might be at the wrong intensity or severity level.” Conversely, “we’re delivering care much more in context and appropriate for each user.”

It seems abundantly important to contemplate the impact of the condition on the population. In a given year, around one in five adults in the U.S. — or 18.5 percent — experiences mental illness, according to the National Alliance of Mental Health.

The National Institute of Mental Health estimates that there were 11.2 million adults at least 18 years of age in the U.S. with serious mental illness in 2017 – accounting for 4.5 percent of all adults in the country.

NAMI showed that broken down by demographic group, the annual prevalence of the condition among U.S. adults looks like this:

  • Non-Hispanic Asian: 14.7 percent
  • Non-Hispanic white: 20.4 percent
  • Non-Hispanic black or African-American: 16.2 percent
  • Non-Hispanic mixed/multiracial: 26.8 percent
  • Hispanic or Latino: 16.9 percent
  • Lesbian, Gay or Bisexual: 37.4 percent

Even in the glare of all that, however,, reported that mental health coverage was being short-shifted by health insurers. In 2015, behavioral care was several times more likely to be provided out-of-network than medical or surgical care. Statistics vary significantly from state to state, the report showed. Forty-five percent of office visits for behavioral health care were out of network in New Jersey; 63 percent in Washington D.C., Milliman stated. Alarmingly, a study published in 2016, at $201 billion, mental disorders paced the list of the costliest conditions in the U.S. in 2013.

Against this backdrop, there are several digital health efforts being addressed toward mental health, among which is the Microsoft-SilverCloud collaboration. Artificial intelligence can help to speed up the understanding and delivery of more personalized mental healthcare. This, in turn, can lead to early interventions that can then improve clinical outcomes.

Cahill, however, was very careful to explain that patient privacy would be safeguarded.

“We’re only looking to utilizing machine learning and the content to ensure the right support for the clinician and coaches by giving them better access to appropriate tools at the right time, while always ensuring privacy.”