A family tragedy was the catalyst for an engineer to create a computer vision and AI-driven platform to warn when senior adults are about to fall

 

“To be able to identify a fall risk, you first have to fall”. This is something that Deepak Gaddipati, Founder and Chief Technology Officer of HealthTech company VirtuSense, knows only too well.

In 2009, Gaddipati’s grandmother fell while walking to the bank and broke her hip. She succumbed to her injury and passed away ten days later. At that moment, Gaddipati dedicated himself to learning from everyone around him, including the housekeepers and several physicians in the family, what could have caused his grandmother’s fall and eventual death. To his astonishment, Gaddipati discovered no one knew his grandmother was at risk of falling, let alone taking appropriate precautionary measures to prevent a tragedy from happening.

Gaddipati quickly realized that most older adults aren’t considered a fall risk until after their first fall, which all too often is tragically, too late. As a technology visionary with clear business insights and expertise in machine vision, deep learning and the Internet of Things (IoT) acquired through his experiences of developing the first commercial full-body, automated scanning system that is widely deployed in many US airports, Gaddipati turned to predictive AI for help.

Gaddipati wanted to prove that technology can proactively identify fall risks in older adults, and that one doesn’t have to wait for a fall to be cared for. To that end, he founded VirtuSense in 2013 and a year later, released the company’s first product, VSTBalance, a predictive fall-risk assessment platform that leverages AI and machine vision to objectively conduct evidence-based evaluations and compares the results to population norms to reliably measure a person’s fall risk. VSTBalance identifies mobility deficits in older adults by examining their risks of falling over a 12-month period based on their balance, gait and function.

The data, coupled with therapy and wellness guidance would offer a granular view into older adults’ mobility as well as clear steps on how to improve. From there, VirtuSense developed its second product, VSTAlert, a round-the-clock remote monitoring platform using AI to detect and alert hospital or facility staff of bed and chair exits before they happen. Normally, staff will only be alerted after the patient or resident has already left their beds, but VSTAlert recognizes a person’s intention 30 to 65 seconds before they physically get up.

According to Gaddipati, VSTBalance has helped communities achieve as much as a 70% reduction in falls and is used in hundreds of sites across the US, including Cleveland Clinic and Johns Hopkins Hospital. As for VSTAlert, it was successfully implemented at John Knox Village, a senior living community in Kansas City where they observed a 79.5% reduction of falls between the one-year period of July 2019 and July 2020, far exceeding their goal of a 27% reduction.

“We focused on falls, especially older patients in hospitals and adult living facilities because there are more than a million falls in hospitals every year,” Gaddipati says. “Right now, human sitters are employed to guard patients from falling, which is expensive and inefficient. We are using AI and computer vision to track everything that’s happening to the bodies of these adults, learn the patterns and use the information to better monitor them. The platform also sends out a warning signal 40 seconds faster than current hospital systems.”

Gaddipati hopes that eventually, VSTAlert will replace bed alarms and extending room monitoring capabilities without the need for someone to continuously watch a video monitoring screen.

In addition, during the COVID-19 pandemic, VirtuSense has also built VSTOne, a continuous remote monitoring and telehealth device helping healthcare providers in acute clinical settings to care for the chronically ill and COVID-19 patients more effectively. The platform accesses patients’ vitals including heart rate, respiratory rate, blood pressure, movement etc, using an array of computer vision and IoT sensors and alerts healthcare workers via a mobile application before patients deteriorate.

This means medical facilities can care for a large volume of patients without scaling staffing or using excess PPE. This not only minimizes the workload for staff but also addresses infrastructure and resources gaps that widen with a sudden influx of patients.

The design of VSTOne was partially inspired by VirtuSense CEO David Park’s father’s sudden death from a blood clot at the age of 71, despite routine doctor’s visits. Both Park and Gaddipati believe, like falling, many other potentially fatal medical conditions are preventable.

“We want to offer clinical insights to our users and their physicians so that they can be safe while living a better and healthier life,” Park says. “Today, we are in hospitals and nursing homes making a real difference in how people age and helping them live longer and more comfortable lives. Tomorrow, our products will play a role in everyone’s lives because we believe AI is the future of healthcare and we are helping to shape it.”