Powered prosthetic devices help amputees walk naturally
Katie Carr, Coordinated Science Laboratory
- ECE PhD student Navid Aghasadeghi uses control theory to help advance prosthetic limbs for lower-limb amputees.
- Aghasadeghi received a $42,232 grant from the National Institutes of Health to pursue his research.
- Aghasadeghi's research could deliver natural locomotion to amputees by establishing an algorithm that uses locomotion trajectories of unimpaired individuals and the amputee's physical characteristics.
According to the Amputee Coalition of America, there are approximately 1.7 million people in the United States living with limb loss and over 1 million lower-limb amputees. The number of lower-limb amputees is expected to double by 2050, especially due to the increasing prevalence of diabetes. Complications of diabetes can cause poor circulation and nerve damage, leading to amputation.
ECE PhD student Navid Aghasadeghi recognized this problem and is putting his control theory knowledge to use to help advance prosthetic limbs, specifically for lower-limb amputees. In February, he received a $42,232 grant from the National Institutes of Health to pursue this research. Illinois Coordinated Science Laboratory and Assistant Professor of Aerospace Engineering Tim Bretl and Northwestern University Associate Professor of Biomedical Engineering and Physical Medicine and Rehabilitation Eric Perreault are advising Aghasadeghi in this research.
“I see health problems as one of the biggest challenges that the world faces,” Aghasadeghi said. “I thought that the knowledge that I had from control theory should be applicable to some of these problems. I wanted to see a complete loop in developing the theory and then testing the theory in application to see if the whole system could work.”
Previously, prosthetic devices were passive, meaning that they behaved like a spring system that didn’t provide power to the individual, but technology has advanced so that there are many powered devices currently available. Powered devices include a motor at the two degrees of freedom – the ankle and knee.
“With passive devices, it has been shown that amputees spend more energy over a gait cycle and that the gait is not natural,” Aghasadeghi said. “With powered devices, if we can provide the right control to the device, we should be able to provide natural locomotion to the amputee.”
The challenge comes with learning how to control these devices and how to provide power to them. Currently, when fitting an amputee with a prosthetic device, a clinician is required to select between a myriad of parameters that match that specific amputee. Aghasadeghi is working to automate this process by using locomotion trajectories from unimpaired individuals and the physical characteristics of the amputee, such as height and weight, to customize controller parameters for the devices.
“Every amputee has to go into a clinic where the clinician chooses the parameters,” Aghasadeghi said. “They would walk with the device, try out the gait and tweak the parameters. We’re trying to customize the controller parameters for the amputees, so clinicians don’t have to spend hours figuring out what the parameters should be. They can just use our algorithm.”
Aghasadeghi is working with the Rehabilitation Institute of Chicago (RIC), which is closely associated with Northwestern University in Chicago, to develop his idea. He spent two summers at RIC and continues to make trips to the Institute for various experiments.
He began by observing how unimpaired humans walk and attempted to replicate that using a prosthetic device. He is working on a theory called inverse optimal control, which hypothesizes that human motor control (moving an arm or leg, etc.) can be modeled as being optimal with respect to a performance criterion. With this theory, Aghasadeghi determines the performance criteria of an unimpaired human walking and uses those criteria to derive controllers for prosthetic devices.
Currently, testing of Aghasadeghi’s algorithm has been done on two non-amputee subjects walking on flat ground.
“Feedback has been positive, in regards to how comfortable the subject was, if the subject’s gait was natural and symmetrical and that the subject didn’t need to use an overhead harness for support,” Aghasadeghi said.
Aghasadeghi has plans to try the algorithm on amputees in the future, as well as extending it to different locomotion modes, such as walking up stairs or down ramps.
“It’s been both fulfilling and challenging to work with experts in the field of rehabilitation,” Aghasadeghi said. “I try to understand the challenges clinicians face and use my theoretical knowledge to develop a precise mathematical representation of the problem they are facing. With this approach, I hope to help improve clinical practice and address challenges in the medical field.”
A video of the prosthetic in action can be seen here.