UVA Professor’s Research Leads to AI-powered Software and a Start-Up Companycfnewman12@gmail.com
Springbok Analytics co-founders (left to right) Joe Hart, Silvia Blemker and Craig Meyer are evaluating biomechanical data. Photo by Dan Addison / UVA Communications
When analyzing how a muscle will respond to injury or disease, scientists often turn to a set of computer models that simulate muscle function. About three years ago, University of Virginia biomedical engineering professor Silvia Salinas Blemker realized a key flaw in those widely used models.
They were based almost entirely on the male body.
“I had been becoming more aware of disparities in data that has been collected over the years, especially in biomechanics and exercise physiology,” Blemker recalled, noting that she does not want to dismiss that work and its importance, but simply to acknowledge the facts. “A lot of studies historically collected data from only male subjects as a way to control for a variable.”
Those studies fueled models that have become ubiquitous in biomechanics, where researchers rely on computer models to simulate muscle function as people move. The models form the basis of all sorts of research and even treatment plans and yet, at most, they could only be scaled for the female body, rather than based on data from real women.
“It was essentially assuming that a woman is a small man,” Blemker said. “Now, of course we realize that is not true, and that there are many differences in musculoskeletal anatomy and structure between men and women.”
For example, Blemker said, women are more likely to experience ACL tears in part because of differences in their musculoskeletal structure, while men are more likely to experience hamstring injuries. Models that account for those differences can help providers better understand how injury occurs and how to precisely treat injuries in both men and women.
“It is so important to get the size and shape of a model right, both with gender and with a range of body heights and masses to represent how the body differs across body mass indexes and how that differs between men and women,” Blemker said.
There was a gap to be filled, and Blemker and her team decided to fill it. She already had many of the tools she needed.
In 2013, Blemker had founded a startup, Springbok Analytics, with co-founders Craig Meyer, a UVA professor in biomedical engineering and radiology and medical imaging, and Joe Hart, at the time a professor of kinesiology at UVA and now a Distinguished Professor and vice chair for research at the University of North Carolina, Chapel Hill.
The startup is based on a first-to-market technology that Blemker, Meyer, Hart and their team developed — an artificial intelligence-powered software platform that turns magnetic resonance imaging, or MRI, scans into 3D analyses of muscle. The scans are faster than typical MRI imaging — about 15 minutes for the lower body — and can be turned into personalized 3D data reports within 24 hours, much quicker than the 60-hour turnaround time typical in the industry. That data can then be used to build the kinds of musculoskeletal models that Blemker recognized were missing – models that accurately represent women as well as men and that can include a wide range of BMIs and body types. In turn, those models contribute to greater diversity and accuracy in the field, giving providers the precision they need to give the best care to all of their patients.
Springbok received encouragement from UVA-Coulter Translational Partnership, the School of Engineering and other resources at UVA to turn their research into a commercial venture, something that the UVA Licensing and Ventures Group now helps facilitate. The technology went to market in 2019 after several years of research and development, with Scott Magaree as CEO and Blemker as chief scientific officer.
Demand for the start-up’s services soared, attracting clients from various industries such as health systems, sport performance and recovery specialists, and universities, among others. In some cases, sport organizations use Springbok’s technology to help athletes rehab injuries and make decisions about when to return to play. In other cases, research organizations are interested in how Springbok’s precise imaging can help them understand how injury occurs or how it can best be treated.
In 2021, for example, Springbok and Blemker were named a co-lead on a $4 million National Football League research project studying hamstring injuries — the most common injury suffered by NFL players. The firm is also developing partnerships with professional sports organizations in women’s soccer and basketball, as well as in some Olympic sports.
“It has been exciting to see research ideas translated to the outside world,” Blemker said.
And there’s more coming.
In particular, she is excited about applying Springbok’s technology to other muscle groups and disorders. The company is preparing to launch a version of its product focused on shoulder injuries and has also developed multiple partnerships supporting clinical research studies in muscular dystrophy, which causes progressive muscular degeneration.
Currently, Blemker is balancing her commitments at Springbok with her continuing research at UVA, where she runs the Multiscale Muscle Mechanophysiology (M3) lab, which focuses on applying a more precise understanding of muscular function to the treatment of diseases like muscular dystrophy and cerebral palsy, or to damage caused by aging. She is also enjoying the ways her two ventures — commercial and academic — work together.
“In my lab, we have a big project doing MRI scans of a large cohort of male and female patients, pairing the scans with functional measurements, integrating those into specific subject models of each person and then creating state-of-the-art female and male [computer] models that can appropriately scale across body types,” she said. “My lab has contracts with Springbok to do the MRI data processing work.”
That kind of collaboration is not unusual at UVA, where many researchers strive to develop innovations that have real-world applications and, in some cases, turn their research into startups, relying on the Licensing and Ventures Group for assistance with the legal and patent process.
Because of particular strengths in engineering and medical research at UVA, Charlottesville has become a biotechnology hub for the region. Biomedical engineering professor Kimberly Kelly, for example, founded ZielBio, which has developed an antibody that could help treat pancreatic cancer.
“It is really nice to have peers at UVA who are doing similar work,” Blemker said. “To some degree, entrepreneurship is not the normal thing to do as a professor, and not something I really imagined doing before coming to UVA.”