A University of Wyoming doctoral student contributed to the creation of an application that allows physical therapists to monitor their patients’ recovery in real time and from a remote location. As a result of his computing innovation, Raghu Raj Prassana Kumar earned an all-expenses-paid trip to the Supercomputing 13 Conference Nov. 17-22 in Denver, Colo.
Prassana Kumar, a UW doctoral student in electrical and computer engineering[BROKEN LINK] (with a focus on parallel computing), won the student poster presentation at the recent Front Range High-Performance Computing (HPC) Symposium. It was the first time UW hosted the symposium.
His poster presentation, titled “Real Time Motion Training for Medical Applications,” focuses on creating a real-time computer model application for physical therapy. The application allows a physical therapist to remotely view, from a computer camera, a virtual model of a patient moving his or her arm during recovery exercises at home.
Using the application, the physical therapist can watch the patient’s range of motion, as well as velocity of movement, to determine if the exercises are executed properly and to monitor progress. The therapist can observe the arm in a horizontal position; flexed or curled inward; and extended.
“When performing an elbow exercise, a patient is expected to maintain the arm right at the horizontal position,” says Prassana Kumar, who is originally from Bengaluru, India.
Normal computers cannot be used to process a high frames per second video containing high-resolution images. As a result, Prassana Kumar says they used “a GPU (graphic processing unit) card as a parallel processor.”
A physical therapist needs a high-definition (HD) video with at least 60 frames per second. Using the GPU card, Prassana Kumar says he was able to process HD images at approximately 10 microseconds per HD frame, providing a theoretical processing capability of 100,000 frames per second. While the application’s speed was increased, the physical therapist now can see body movement in three dimensions. The application’s algorithm is actually built for 3-D analysis.
“A basic model representing a hand is ready. We are working on building the full hand,” Prassana Kumar says. “What we’re trying to do is get this project on the 3-D CAVE.”
The 3-D CAVE (Cave Automatic Virtual Environment) is the centerpiece of the Shell 3-D Visualization Lab in UW’s Energy Innovation Center.
Prassana Kumar’s poster presentation was chosen the winner over six others. The symposium competition required each presenter to explain the basic concepts of his or her project within four minutes. Additionally, each presenter had one hour to answer questions asked by high-performance computing officials.
“We had to talk to the audience for 60 minutes,” Prassana Kumar says. “The judges were there, but I did not know who they were.”
Prassana Kumar worked on the real-time model application with Suresh Muknahallipatna, a UW professor of electrical and computer engineering, who serves as Prassana Kumar’s Ph.D. chair; and John McInroy, a UW professor and electrical and computer engineering department head. McInroy is Prassana Kumar’s Ph.D. co-chair.
McInroy developed the computer vision algorithms. Muknahallipatna and Prassana Kumar modified the algorithms for real-time model implementation on GPUs. All three later worked to create the remote physical therapy application.
“Raghu’s poster win highlights the need for addressing the interactions between the hardware and software aspects of GPU programming to achieve efficient high-performance computing,” Muknahallipatna says. “Application of the GPU-based HPC for remote physical therapy to provide real-time feedback is unique.”
Prassana Kumar says he is excited at the prospect of meeting potential business contacts and learning the latest in parallel computing at the Supercomputing 13 Conference. He attended the 2011 version of the Front Range HPC Conference at Colorado School of Mines.
“Usually, at these (supercomputing) conferences, a lot of experts give their keynote talks,” says Prassana Kumar. “This is very important for Ph.D. students to learn what’s happening in parallel computing at the industry level and how they solve problems. At the industry level, each problem is addressed uniquely, and globally.”
After he completes his doctoral dissertation later this year, Prassana Kumar hopes to secure work in an industry lab, such as at Intel, Microsoft or Nvidia.
“If I am able to get into a lab that facilitates parallel computing, that would be really good,” he says.
