USF researchers developing AI-powered knee sleeve to help prevent ACL injuries

TAMPA, Fla. - Researchers at the University of South Florida (USF) are working to develop a high-tech knee sleeve that uses artificial intelligence (AI) to identify movement patterns that could lead to ACL tears, offering future protection to athletes. 

How it works

What we know:

At first glance, the prototype looks like a typical knee sleeve. But this wearable is powered by AI.  

By monitoring an athlete's movements in real-time, it can identify risky movement patterns — like an awkward pivot or an improper landing — before the ligament snaps. Then, the sleeve would deliver immediate feedback to the athletes — either by vibration, sound, or some sort of notification on an app. 

"The idea is to use artificial intelligence and a wearable that can then actually interact with the individual and train them in real time while they're practicing," said Dr. Nathan Schilaty, associate professor in the USF Health Morsani College of Medicine Department of Neurosurgery, Brain and Spine and USF Department of Medical Engineering. "So, it's like you have an athletic trainer attached to your knee, right? Telling you what you need to do and how to change your biomechanical behavior so you avoid that injury."

The non-contact danger

By the numbers:

Schilaty said about 80% of ACL injuries happen in non-contact incidents, such as sudden stops, improper landings, and awkward pivots. Schilaty notes that young girls, specifically between the ages of 15 and 19, are at a particularly high risk of injury. 

Additionally, corrective surgeries can be expensive, costing between $20,000 and $40,000, according to USF. And the recovery process from an ACL injury can be long and difficult. 

Just ask Tampa Bay Buccaneers outside linebacker David Walker, who tore his ACL during training camp last year and missed his entire rookie season. But he's finally back on the field — even getting mic'd up during the preseason game last week.

Predicting injury vs. evaluating injury risk 

What they're saying:

Ultimately, Schilaty says the goal is not to predict an injury with absolute certainty but rather to assess risk and steer athletes away from the danger zone.

"We know there's certain biomechanics when you're landing and when you are accelerating or cutting and twisting that are going to be something that puts you closer to that threshold of injury," Schilaty said. "So the idea is to just take people as far away from that as possible so that they don't get the injury."

What's next:

Moving forward, the USF researchers are working to secure additional funding to expand testing to eventually include clinical trials with live patients.

"Of course, one of the things we need to look at is, when we deliver a signal for modification of that behavior, what are people going to modify?" Shilaty said. "You know, when you leave someone to their own auspices to change their behavior, hopefully they change it the correct way and not the wrong way, right? But it should be then they're getting fewer notifications because that means they're moving into a better scenario."