This Startup Spots Stress in Real-Time to Help Prevent Depression and Other Conditions

Illustration of an IMU sensor with stressful faces floating above it
Illustration: iStockphoto/IEEE Spectrum

By: Kathy Pretz

THE INSTITUTE By any measurement, 2020 has been stressful for just about everyone because of the COVID-19 pandemic. Fear about the virus and concerns about our health and that of loved ones can be overwhelming. Add that to the other tensions many of us have at work, at home, and at school.

When a person is stressed enough, the fight-or-flight response kicks in. The sympathetic nervous system causes a sudden release of hormones—which increases heart rate, blood pressure, and perspiration.

The first step in controlling stress is to know its symptoms, but because most people are used to some stress, they don’t realize how bad things have gotten until they reach a breaking point. Over time they could experience serious health problems such as heart disease, high blood pressure, and diabetes, as well as depression and other mental health woes, according to the U.S. National Institute of Mental Health. More than 264 million people suffer from depression, the World Health Organization reports.

What if there was a way to measure in real time when a person was becoming stressed, so the condition could be managed immediately using evidence-based methods? That’s the idea behind Philia Labs, a startup in Melbourne, Australia, that has developed a platform with a wearable device designed to measure physiological stress indicators.

The product is aimed at health care providers and mental health professionals, as well as people who want to monitor their own stress level.

“We are quantifying stress in the body in real time,” says Dilpreet Buxi, the startup’s cofounder and chief executive. “The hardware platform and software will enable interventions both through a health care provider and by the patient to basically enable better health outcomes and a better quality of life.”

STRESS INDICATORS

To confirm whether someone suffers from stress, Buxi says, doctors typically use a questionnaire such as the Kessler Psychological Distress Scale or the Depression Anxiety Stress Scales. Such forms help assess a person’s emotional state and quality of life based on situations that might trigger anxiety. But because they are self-evaluations, the results can be inaccurate.

Some of today’s fitness wearables claim to measure stress. They use data about heart rate, sleep, and level of activity to infer how stress is affecting the wearer. But, Buxi says, the results from such devices haven’t been clinically validated.

In contrast, he says, Philia aims to measure physiological data that has been shown to more closely align with stress response and to pursue focused clinical testing. Philia’s wearable, which is worn on the wrist for at least six months, uses optical sensors to measure heart rate and blood flow. Electrodes measure “galvanic skin response”—changes in moisture caused by sweat-gland activity that can indicate a person’s emotional state, Buxi says.

Galvanic skin response refers to the electrical conductivity of the skin,” he says. “In other words, when you break out into a nervous sweat, the electrical conductivity will change.”

Philia will initially pilot its technology on patients undergoing depression treatment, he says, adding that a clinician will prescribe the device and a clinical monitoring program for the patient. Physiological and self-reported data are captured from the patient’s sympathetic arousal—that fight-or-flight response—and computed. Trends in sympathetic arousal activity over weeks and months are calculated to determine whether a patient requires an intervention such as a change of medication or psychosocial treatment. All the information is stored in the cloud.

For patients who previously have had depression, early intervention could help reduce the risk of a recurrence, Buxi says.

“According to our conversations with psychiatrists,” he says, “stress that results in sympathetic arousal is a leading cause of relapse and needs to be monitored in order for the psychiatrist to intervene earlier.”

He says the likelihood that a person who has recovered from depression will relapse in the first year when suffering from stress is 20 percent to 50 percent.

“The platform will enable the medical provider to make better decisions,” he says. For patients, he adds, “the goal is to basically help them adopt better techniques for stress management.”

Philia has several partners including medical institutions and research universities. It is running pilot programs with 11 health care and wellness organizations. The company has filed a provisional patent application.

The startup has a proof-of-concept prototype for the wearable, which is built using off-the-shelf parts and is moving to a minimum viable product that will be used after a study and several trial programs are completed next year, Buxi says. A lab study on 60 patients is currently happening and will end in April. A small trial on those with mild depression patients starts in January, and a multi-site trial in depression relapse will begin in June. He says the trial is with a corporate health provider, which can expand the company’s market portfolio to non-clinicians.

The company will be seeking regulatory approval for the platform after it undergoes clinical trials.

INSPIRATION

A biomedical engineer, Buxi worked from 2008 to 2012 at the Holst research center in Eindhoven, the Netherlands, where he integrated state-of-the-art technologies for wearable health care devices. After that, he relocated with his family to Australia, where he pursued a Ph.D. at Monash University in Melbourne. For his research-project thesis, he developed a wearable blood pressure monitoring system based on pulse transit time—for which he was granted a patent from the Australian government.

Several of his research papers are published in the IEEE Xplore Digital Library.

Buxi got to thinking whether he might apply his Ph.D. work to the problem of measuring stress.

He began working on the idea in 2017 as a side project, and in 2018 he formed a proprietary limited partnership with the startup’s cofounder, Alexander Senior. Today the company has seven employees—a mix of engineers, scientists, and entrepreneurs. The company also has collaborators from industry and academia who have expertise in machine learning, biomedical machine learning, and physiology.

The business has largely been funded by a venture capitalist and is close to completing its seed funding round.

LEARNING TO BE AN ENTREPRENEUR

Buxi says his biggest challenge was transitioning from being an engineer and scientist to becoming an entrepreneur.

“You need to think in terms of what is the problem you’re solving that requires a solution that somebody is going to pay money for,” he says. “That’s completely different from doing an investigation in the lab.” As an entrepreneur, “you have to find a solution where you can repeatedly get new and old customers to pay [so that you have] new and recurring revenue.

“That took a lot of learning,” he says. “In fact, even today, I think more commercially, but I’m still pretty academic. And sometimes it shows.”

He says he got help with how to run a startup from IEEE’s Founder Office Hours program, which seeks to assist early- and growth-stage technology entrepreneurs from the IEEE community. It connects entrepreneurs to mentors who can provide feedback and potentially help them grow their company.

In a testimonial about the program, Buxi says he got assistance with validating the product, thinking about the pros and cons of various business models, and refining an intellectual-property strategy to create value.

“The program shaped our thinking a bit,” he says, “to make our approach more practical.”

This article originally appeared in IEEE Spectrum on 30 October 2020.