Vital Audio Improves Access to Vital Signs

For Nyamitse-Calvin Mahinda, a college job as a patient representative for a health care company started him on his entrepreneurial road by showing him the complexity of the American health care system. But it was his mother’s challenges with the healthcare system before her death, and his passion to make remote access to care more equitable, that inspired him to create Vital Audio.

Born and raised in Tanzania, Mahinda moved to the United States via Germany more than a decade ago to continue his education. His interest in biology led him to Lincoln University in Missouri, where he completed a bachelor’s degree in science as a pre-med student.

“When I was in school, I started working in the healthcare space as a patient access representative in the billing department, and then moved into more clinicals roles, eventually as an EMT at several clinics,” he said. “That helped me to understand all the levers that make healthcare system work.”

Mahinda’s mother was diagnosed with cancer after she also moved to the U.S. While he was still in Missouri, she would call with questions. “A single call goes a long way,” said Mahinda. “That all falls under patient after-care. Once she was in remission, care gets fragmented, and it’s harder to keep up. I realized that wrap-around care should be part of the picture for every illness.”

“My mom called one night, and I wasn’t able to take her call—I was working in the clinic,” added Mahinda. “She died later that night, before getting the care she needed. A death should never result from the complexity of navigating care, or a lack of care. I wondered how we could prevent this from happening to someone else.”

These challenges were top of mind for Mahinda when the pandemic hit and he began working for Cityblock Health, a tech-enabled virtual primary care service. “I was collaborating with physicians to create a process for implementing new technologies for remote care. I saw firsthand the challenges of the tech—that it was costly, or consumers didn’t know how to use it.”

Eventually he entered NYU, where he is now completing a master’s in biomedical engineering, focusing on interactive medical robotics and digital signal processing. “My mindset had changed,” said Mahinda. “I thought I had to be patient-facing to serve patients, but I learned there’s more than one way to do that. I wanted to work on broadening accessibility and alleviating the technical and confusing elements that limit it.”

That goal led to the concept behind Vital Audio, which seeks to help remote patients provide physical health care data such as vital signs to medical providers through traditional phone calls. Vital Audio uses a ten-second speech sample for quick assessment, using voice to get data such aa s heart rate and rhythm regularity, and then reports it to the provider. 

Mahinda met his co-founder, Harsh Sonthalia, through NYU’s Future Labs incubator center. Sonthalia, originally from India, had come to NYU to pursue a master’s in computer science. He was introduced to Mahinda, who had conceived of Vital Audio and was looking for someone who could code.

Mahinda and Sonthalia first participated in I-Corps while at NYU; Mahinda’s lab connected them with the NYU School of Engineering venture arm. “In our first customer discovery, I spoke with about 200 doctors and others in the field,” Mahinda said. “We did a lot of hypothesis formulation, validating and invalidating, that allowed up build strategically in the process rather than building it and taking to customer and then having to make changes.” Once Mahinda had pulled together a small team, they participated in a national I-Corps cohort, based out of the southwest. 

Sonthalia described the significant learnings they gleaned from that I-Corps cohort. “One of the biggest takeaways was gaining a better understanding of what people thought of it, whether it would be easy or difficult to adapt into their daily use and workflow,” he said.  

Like many teams that go through I-Corps, Vital Audio had to pivot. “Our initial focus was primary care,” said Mahinda. “Today, we are focusing on specialty care cardiac and pulmonology, and plan to continue expanding to encompass remote care delivery both pre- and post- hospitalization.”

Vital Audio’s software uses a series of audio filtering techniques to remove frequencies that do not correlate with physiological data extraction. It then visualizes the frequencies using a Short-time Fourier-transform, which is used to determine the content of a signal as it changes over time. It finds patterns that match with the desired physiological data extraction—in this case, human vitals. Once the software has isolated the desired information, it can calculate the vital signs.

“Our idea is that your cell phone becomes your medical device,” said Mahinda. “This would open up more equitable access to care, especially remote care, and reduce the limits on patient tracking. For example, right now patients must rely on wearables, which cost less, but only 10 percent of what they capture is actionable data. The provider or patient should not have cumbersome steps to get or share data and support patient care.”

“My first exposure to health care tech was during a semester of study I did at MIT, where we collaborated with Harvard Medical School,” said Sonthalia. “This area was already something I was interested in. When Mahinda presented his idea, I thought this was interesting but also insane—no one had ever done this before. And I figured if we could pull it off, it would be great.”

Mahinda cites his biggest takeaway from his I-Corps experience as something that continues to infuse his approach to growing Vital Audio. “I always go back to the customer,” he said. “The customer is the center of the building process. The tech must provide a service and alleviate pain or it’s not worthwhile. That is my guiding principle—put the customer at the forefront of discovery and a business alignment.”

Mahinda has also learned from numerous resources offered as part of NYU’s Tech Venture Program. After participating in the first I-Corps program, TVP connected him with the NYU School of Engineering venture arm, NYU Tandon Future Labs. Mahinda also received startup coaching and mentorship and participated in the NYU Entrepreneurs Festival. “We’ve benefited from strong mentorship, industry coaching, and collaborations that facilitate our ability to execute on so many levels in such a short time,” added Mahinda.

Along with Mahinda and Sonthalia, other team members include an expert in audio signal processing and two clinicians who Mahinda met at Cityblock. The team has discovered some major insights, demonstrating that the software leads to a 20-30 percent reduction in triage and administrative time, better allocation of nursing resources to critical tasks, and potential review increases from increased capacity for providers to see more patients.

The Vital Audio prototype is up and running on a HIPAA-certified platform, so it is data privacy and security compliant. “All patients need to do is know how to operate their part of the phone call; the software is on the health care side. When they answer the call, they repeat sentences or hold vowel sounds.” 

Mahinda described one of Vital Audio’s biggest challenges as inherent in offering a novel technology. “Some people jump on the tech right away, as they see opportunities in remote care and after care,” he said. “But it’s a curve ball to market something new.”

“Also, because this hasn’t been done before, there’s no prior study or starting point of data or precedent to leverage,” added Sonthalia. “We have to make our own data set from scratch, our own study from scratch. But it’s been fun to pursue that challenge.”