Editor’s note: This is one of five profiles of finalists for NVIDIA’s 2017 Global Impact Award, which provides $150,000 to researchers using NVIDIA technology for groundbreaking work that addresses social, humanitarian and environmental problems.
Duane Storti has a bone to pick — 20 of them, in fact.
Storti, a professor at the University of Washington, works with the Center for Limb Loss Prevention and Prosthetic Engineering at VA Puget Sound to study how the 20 bones that make up our feet work together while walking.
The research for this can be slow due to the time it takes to process data. But Storti and his collaborators are changing that with a faster, GPU-powered approach to track and scan the lower limbs in a patient.
Using both 2D and 3D X-ray imaging, the team can study the shape of each foot bone, and understand how the bones move together when walking. With this new method, they hope to improve how effectively doctors compare lower limb treatments in veterans.
Their work has placed them among among five finalists for NVIDIA’s 2017 Global Impact Award. Our annual grant program totaling $150,000 goes to researchers using NVIDIA technology for groundbreaking work that addresses social, humanitarian and environmental problems.
These Bones Were Made For Walking
Clinical trials for foot-bone treatments need a substantial number of patients matched with treatments that best suit their ailments. If it takes a long time to process patient imaging data, however, optimal matches can be delayed. And this can render clinical trials infeasible, as not enough patients are able to participate.
In the image restoration process, which plays a crucial role in treatment evaluation, the numbers add up fast.
“You’ve got 20 bones (in your foot), each of which has to be imaged in about 100 different configurations until you finally get a reasonable match, and then you’ve got several hundred frames of the movie,” Storti said.
Registering a full set of walking data could take a week and a half. And studies require a few dozen patients for each treatment being evaluated. Plus, doctors are typically comparing two or three possible treatments for a particular disorder.
With Storti’s CUDA-powered method, clinical trials for treatment studies became possible as the computing time was reduced from a week-and-a-half to one day.
“Aligning the bone shapes with the X-ray pictures involves doing a lot of math, and we use powerful GPUs to provide the boost in computing power that we need to do all that math in a reasonable amount of time,” Storti said.
A Step Forward in Integrating GPUs in Healthcare
While Storti’s current research and new algorithm is used to study problems in feet, he foresees that his work will assist future research of other body limb ailments, including the knee and spinal cord.
With AI advancing into more industries, Storti hopes that his research is just the beginning of integration of GPUs in healthcare.
“To be able to bring the CUDA-scale parallel computing to (healthcare), that’s something that you don’t have to have a national laboratory for,” Storti said. “If you just have a good laptop, you can do meaningful things.”
The winner of the 2017 Global Impact Award will be revealed at the GPU Technology Conference, May 8-11, in Silicon Valley. To register for the conference, visit our GTC registration page.
Other Global Impact Award 2017 finalists include:
Check out the work of last year’s Global Impact Award winner.