Startup reduces unnecessary spine surgeries using AI
website – https://www.theseus.ai
Newly formed Theseus AI plans to reduce unnecessary back surgery referrals by providing more objective MRI data to physicians.
A Los Angeles-based startup recently struck a licensing deal with the University of California- Los Angeles to create a tool that can analyze spine images to help decide whether patients should be a candidate for surgery.
Theseus AI was spun out of research conducted by Dr. Luke Macyszyn, a neurosurgeon at the UCLA spine center. The startup struck an agreement to license the technology from UCLA in September.
EO Sam Elhag presented Theseus’ work in front of a group of colleagues at the SoCalBio Digital Health conference on Friday. The company is addressing a costly medical problem: more than 250,000 spine surgeries are preformed annually, Elhag said, and a significant portion of them might be completely unnecessary.
Studies show a substantial number of back surgeries fail to relieve pain. Estimates range between 20 percent and 40 percent of surgeries fail, according to a 2016 review by the Journal of Pain Research.
“This is somewhat personal for me in that my own father had spine surgery twice,” Elhag said. “To this date, it’s uncertain as to whether or not it all made sense.”
Elhag’s solution is to make the interpretation of spine MRIs less subjective. Theseus uses an algorithm developed by Macyszyn to provide objective measurements of spinal stenosis, or the narrowing of spaces within the spine, by analyzing MRI images. The company will also compare those measurements to patients of the same gender, age and height.
MRI reports use terms such as mild, moderate or severe to describe the condition, which are undefined and can often fluctuate.
“If you give the same physician the same MRI on different days they may give you a different diagnosis,” Elhag said. “We need to improve that by giving those folks more objective data.”
Elhag is a technologist by background. He previously founded Engrade, an education software company that was acquired by McGraw-Hill Education in 2014. Years prior, he helped build a strapless heart rate algorithm that was later licensed to EA Sports and Microsoft.
Still in its early stages, the startup has developed the first version of its product and filed for initial patent applications. The company plans to seek regulatory approval through the Food and Drug Administration’s 510(k) pathway, and has identified a comparable predicate device. Theseus also raised $125,000 this year in a friends-and-family round.
In the long term, Elhag hopes Theseus’ work with university health systems will allow the company to make inroads with payers, reducing the overall number of unnecessary surgeries. He also hopes it will help patients reach the right treatment faster, instead of bouncing between insurance companies, primary care doctors and surgeons.
“Back pain is a big problem. The thing we’re aiming to solve is very specific: the fact two in 10 folks that meet with a surgeon end up getting surgery,” Elhag said. “In general, folks are meeting with a surgeon even though they don’t need to have surgery.”