Mental health startup uses speech biomarkers to detect depression and anxiety

Kintsugi’s AI software analyzes the human voice to detect depression and anxiety, providing clinicians with “a more complete, 360-degree view of the patient” that is sometimes lost in virtual care.


“Tthe quick brown fox jumps over the lazy dog,” Rima Seiilova-Olson says slowly and emphatically over Zoom.

The simple sentence is of immense value to mental health, she explains, smiling as if to acknowledge that it may not be so obvious how a silly sentence could be so meaningful to a computer programmer and leader of an artificial intelligence startup. .

The short saying includes every letter of the alphabet and phoneme in the English language, says Seiilova-Olson, an immigrant from Kazakhstan who is the co-founder and chief scientist of Kintsugi Mindful Wellness. Kintsugi believes these sounds provide invaluable insights that can help caregivers better support people with depression and anxiety.

The Bay Area-based company is building AI software that analyzes short speech fragments to detect depression and anxiety. This voice biomarker software is being integrated into clinical call centers, telehealth services and remote monitoring apps to screen and triage patients requesting support, making it faster and easier for caregivers to assess their needs and respond.

“There’s just not much visibility into who is severely depressed or anxious.”

Kintsugi CEO and Co-Founder Grace Chang

Seiilova-Olson, 36, first met co-founder and CEO Grace Chang, 40, a Taiwanese immigrant who now lives in Berkeley, in 2019 at an open AI hackathon in San Francisco. Surprised to cross paths at a male-dominated event, the women began comparing notes about their respective personal challenges in trying to access mental health care: Seiilova-Olson struggled to find a therapist during postpartum depression with her first child, and when Chang needed her own support, she said it took months for someone from Kaiser to call her back.

“If you live in the Bay Area, you can push a button and a car can come to you or food can come to you,” Chang says. “But this was a real challenge.”

As engineers, they viewed the dilemma differently from clinicians.

“We saw this as an infrastructure issue, with so many people trying to jam through that front door,” Chang explains. “But there’s just not much visibility on who is severely depressed or anxious, who is low to moderate. And if we could provide this information to those primary care practitioners, then we might have a chance to significantly alleviate that bottleneck.”

Kintsugi was born from that idea in 2019. It sits in a competitive space of health technology startups such as Ellipsis Health and Winter Light Labs, which use speech biomarkers to detect mental health or cognitive problems, based on research showing that certain linguistic patterns and features of the voice of a person can be correlated with psychiatric or neurological disorders. Kintsugi raised $8 million in seed funding last year under the leadership of Acrew Capital, and announced in February that it had closed a $20 million Series A round led by Insight Partners, which PitchBook said the company had at nearly $85 million. estimated.

Personal health care facilities typically use questionnaires to measure the severity of patients’ anxiety or depression, measures known as PHQ-9 and GAD-7 scores. But during telehealth visits or telephone consultations — where face-to-face interaction is lost, making it harder to catch symptoms — Kintsugi’s technology helps fill that gap.

Nicha Cumberbatch, deputy director of public health at Spora Health, a provider focused on health equity and people of color, uses Kintsugi’s software to assess women in her fully virtual doula-led maternal health program, Spora Mommas. The voice analysis tool, which Spora started using for patient consultations a few weeks ago, has helped Cumberbatch identify women who have or may develop anxiety and depression before, during or after their pregnancy. When a patient on Zoom starts talking to a Spora doctor or doula, Kintsugi’s AI begins to listen to her voice and analyze her. After processing speech for 20 seconds, the AI ​​spits out the patient’s PHQ-9 and GAD-7. The employee can then use that mental health score to decide what additional tests may be needed and how best to advise or refer the patient to resources, such as a psychiatrist, cognitive behavioral therapist, or obstetrician.

Cumberbatch says Kintsugi’s technology allows her to “monitor her patients more closely” and then move forward with proactive recommendations to reduce their symptoms. And while it’s not intended to replace clinicians or formal medical evaluations, she adds, it could be used as a screening tool to “allow us to have a more complete, 360-degree view of the patient when we’re not seeing them.” in our face.”

“That technology… [allows] us to have a more complete, 360-degree view of the patient when we don’t have them in front of us.”

Nicha Cumberbatch, Deputy Director of Public Health at Spora Health

dr. Jaskanwal Deep Singh Sara, a Mayo Clinic cardiologist who has collaborated with Ellipsis and led research into possible use of voice biomarkers in cardiology, warns that while the technology holds promise for healthcare, the field still has a long way to go. ensuring that it is accurate, safe and beneficial for both patients and clinicians.

“It’s far from ready for primetime,” says Dr. Sara. Studies in psychiatry, neurology, cardiology and other fields have shown a link between voice biomarkers and various conditions or diseases, but they haven’t shown how this relationship could be used to improve clinical outcomes, he says. Such research is “not the same as saying, ‘How can we instrumentalize it in clinical practice, and how feasible is it? How effective is it in measuring a person’s medical trajectory?’, he explains. “If it doesn’t bring benefits in the way we manage them, then the question is, why would you do it?”

He says addressing those questions is “one of many next steps we need to take on this” and that larger clinical trials are needed to answer them. “If it makes healthcare cheaper or more efficient, or if it improves patient outcomes, that’s great,” he adds. “But I think we need to demonstrate that with clinical trials first, and that hasn’t happened.”

To address these issues and validate its software, Kintsugi is conducting clinical trials, including with the University of Arkansas for Medical Sciences, and the National Science Foundation has awarded Kintsugi multiple grants to accelerate its research. The company is also pursuing FDA “de novo” approval and continues to build its own dataset to improve its machine learning models. (Data and insights from Kintsugi’s voice-journaling app, as well as calls to call centers or telehealth providers and clinical collaborations with several hospitals, are all becoming part of a huge dataset that powers Kintsugi’s AI.) Seiilova-Olson says this is self-generated, untethered proprietary dataset is what sets Kintsugi apart in AI healthcare – where many technologies rely on external data from electronic health records.

That collection of data on individuals’ speech can be worrisome, especially in mental health and wellness, which is widely regarded as a regulatory Wild West. (These products and services are often not subject to the same laws and strict standards that govern how licensed clinicians provide formal medical care to patients.) But Kintsugi’s founders say patient privacy is protected because it doesn’t matter to the technology. what people say, but how they say it. Patients are also asked to register their consent and care is not affected by their decision to opt in or out, according to the founders.

Kintsugi says it has helped an estimated 34,000 patients. The company currently operates a major health system with 90 hospitals and clinics in 22 states, and they operate a healthcare management call center that handles approximately 20 million calls per year. It also partners with Pegasystems, which provides customer service tools for healthcare and other industries, to help payers and providers handle incoming calls. Chang says other clients include Fortune 10 enterprise payers, pharmaceutical organizations and digital health applications focused on remote patient monitoring, but she has not yet been able to share their names. Kintsugi’s clinical partners include Children’s Hospital Colorado, Joe DiMaggio Children’s Hospital in Florida, Chelsea and Westminster Hospital in London and SJD Barcelona Children’s Hospital in Spain, Chang said.

Prentice Tom, Kintsugi’s chief medical officer, adds that it is working with the University of Arkansas to explore how the tool could be used to identify patients with suicidal thoughts or increased or severe suicide risk, as well as with Loma Linda University to look into how the technology can be used to detect burnout among clinicians. The team is also looking for ways to expand availability and use for younger and older patients, as well as maternal and postpartum populations. And besides the patients themselves, it’s perhaps the nurses who benefit most from Kintsugi’s work, according to the founding team: with a triage tool that helps reduce administrative work or time spent asking generics. questions, nurses can move patients seamlessly as they travel.

But Tom, a Harvard-trained emergency medicine physician and former faculty member of Stanford University’s Department of Emergency Medicine, says Kintsugi is now doing much more than addressing infrastructure problems. It is democratizing access to mental health care, Tom said, moving away from a physician-centered paradigm more focused on people with depression so severe they need medical evaluation.

“This tool basically creates a picture of mental health in terms of mental well-being,” said Tom, “where everyone has the opportunity to understand where they sit on the spectrum and that basically stratifies treatment options far beyond the current ones.” infrastructure.”

MORE OF AI 50 2022

MORE FROM FORBESAI 50 2022: The best AI companies in North America shaping the future
MORE FROM FORBESAI 50 methodology: how we selected our honorees
MORE FROM FORBESThe $2 Billion Emoji: Hugging Face Wants To Be Launching Pad For A Machine Learning Revolution
MORE FROM FORBESAI Upstart Waabi adds self-driving veterans to race to commercialize robot trucks

Leave a Comment