Article Text
After watching his grandmother suffer repeated fall-related injuries despite reassuring physical therapy checkups, junior Jony Li spent more than five months developing a new way to assess mobility problems.
His project, SmartGait, combines a lightweight virtual reality headset, pressure-sensing insoles and video analysis. The app earned second place in the Congressional App Competition, with the result announced in late December.
Li hopes the system can help patients receive more frequent, affordable assessments while giving medical professionals better information about changes in a person’s gait, or walking pattern.
More than 14 million older adults fall in the United States each year, according to figures cited by Li, and nine million of those falls result in injuries. Mobility and gait problems are often contributing factors.
Li said his grandmother experienced a severe decline in mobility over the previous year. Because his parents were busy working, he frequently accompanied her to physical therapy.
Therapists often said she appeared healthy, he recalled, yet she was rushed to the hospital several times with fall-related injuries only weeks after checkups. That gap helped inspire SmartGait.
The app can be installed on VR headsets and offers customizable settings and walking activities for mobility assessments. Smart insoles placed inside a patient’s shoes measure the distribution of pressure across the feet.
A video-analysis component built with OpenPose examines joint angles and stride patterns to identify potential abnormalities.
Li found that gait analysis is difficult because walking patterns vary widely and can change with fatigue, recovery and surroundings. He also concluded that infrequent clinic visits and costly assessment equipment can make mobility problems harder to detect.
Developing the system required Li to learn about artificial intelligence anomaly-detection models while managing the demands of his junior year. He studied previous mobility research, reviewed different methodologies and costs, and consulted technical professionals and his parents, who work in technology.
He ultimately selected smart insoles and video analysis to measure foot pressure and joint movement. Another major challenge was deciding which measurements mattered most because his computer could process only a limited amount of data at once.
His younger brother, Ethan Li, said Jony worked through repeated debugging problems while trying to balance technical demands with an accessible user experience. Ethan called the second-place finish a major accomplishment.
Early testing with a 71-year-old recovering from surgery indicated that the system was comfortable and intuitive to use during daily activities, Li said. He nevertheless considers SmartGait a work in progress.
Li plans to lower the system’s cost and improve the algorithm’s precision, with the long-term goal of making gait and mobility assessments more convenient for older adults. He said the problem reaches far beyond his grandmother and represents a national health care concern.