Robert Petrea is a bilateral above-knee amputee who lost his legs when he was 25 years old. With the help of prosthetics, he was able to walk for 10 years beginning in 1978. He stopped using his prosthetics in 1988 because they hurt him too much.

“If you used the prosthetic devices as a tool, they weren’t a very good tool. Sitting in a wheelchair was a lot easier, physically,” said Petrea, a principal research specialist in agriculture at the University.

Petrea thinks that if his prosthetics had some kind of padding to prevent the ends of his legs from rubbing the device, he may have continued to use them. Petrea suggests that a cushioning agent could preclude a lot of problems with the device because essentially the end of the leg floats around and rubs on the prosthetic itself.

At the University, Navid Aghasadeghi, a doctoral student in electrical and computer engineering, is developing an algorithm to change the way lower-limb prosthetic devices function with the human body and to quicken the process for fitting the devices more comfortably than ever before.

Traditional lower-limb prosthetic devices need to be tuned to each patient specifically, and the process can take hours and cost a significant amount of money. In current practice, patients go to a clinic for a fitting. Doctors then have to tune many parameters to reach a certain gait, to fit each patient specifically. This entire process can take about four hours and only adjusts the prosthetic for level ground walking. For varying ground levels, the parameters must be adjusted by allowing the patient to walk on slopes and then fine-tuning the device accordingly. Often, patients must go to a clinic several times a year for regular fine tuning and adjustments.

Kevin Carroll, a world-renowned prosthetist, said that comfort is critical in these devices. He helped to develop devices like those that double-amputee sprinter Oscar Pistorius used in the 2012 Olympics. Carroll also developed a gel called Wintersgel, named after a tail-less dolphin named Winter who received a prosthetic tail from Carroll. This same sticky gel is being used in human prosthetics to provide better comfort as well.

“Today, we can use scanning technology to scan limbs and provide more comfort to prosthetics by molding a more accurate fit,” Carroll said.

But even with this cushioning material, Carroll compared the process to trying on shoes: If one pair of shoes is not comfortable, one can simply take them off and try on a new pair. For prosthetics, however, it’s not that simple.

To achieve comfortable walking through the use of prosthetics, Aghasadeghi’s learning algorithm replicates natural walking.

Starting in the summer of 2011, Aghasadeghi contacted researchers, including his current co-adviser Eric Perreault at the Rehabilitation Institute of Chicago and proposed a new way to manipulate the current design of prosthetics and to use robotics to make the devices perform as desired.

His algorithm, called inverse optimal control, captures how humans produce movement and readily adapts the prosthetic to new terrain. The algorithm produces movement for amputees by predicting how a person would walk on given terrain and then adjusting itself to fit the physical attributes of the amputee. Instead of going to the hospital to collect data and tune the prosthetic device, the algorithm collects its own data, such as the patients’ weight and height, and automatically tunes the device for the amputee to walk in different landscapes.

Because current prosthetics do not adapt to new terrain, they lack much of the stability most people are able to achieve with their legs. Basic stability in prosthesis has been a major development, but it’s far from perfect. However, advancement of technology, including Aghasadeghi’s algorithm, provides more stability by customizing the device for amputees with different heights and weights. Carroll also suggests the fear of falling has been drastically reduced because of similar advancements in technology.

Aghasadeghi’s algorithm reduces the amount of energy someone needs to move because traditional devices are not powered. State-of-the-art prostheses overcome this problem, but even those still need to be fine-tuned by a professional.

Further, making lighter prosthetics decreases the energy needed to walk and makes devices easier to manipulate, but a lighter device means less durability.

Aghasadeghi has been testing his algorithm on a device developed at Vanderbilt University. This device combines power with durability, running on rechargeable battery power. This battery can last for a full day’s use, or approximately 8 hours of constant use, before needing recharged. If the batteries die, however, the device will switch to a “safety mode,” which functions like traditional spring-loaded prosthetic devices. Traditional prosthetic devices cause the user to exert more energy than normal walking would require, while the prosthetic that Aghasadeghi is testing provides power to the device to provide a smoother experience.

To achieve this, an interface connected to the nervous system via electrodes, records signals from the muscles. The signals then detect how a person is moving, and the prosthetics adjust.

There are two interfaces between the prosthetic device and the patient, Aghasadeghi said. The electrodes, attached to the skin, read both the electrical signals from the muscles and the interaction of the muscles with the device. With that information, the device uses Aghasadeghi’s algorithm to respond to the patient’s movement.

Most recently, Aghasadeghi has tested his algorithm with prosthetics on level ground and hopes to continue testing it on amputees on different terrains in the near future.

“Engineering gives us a lot of the tools to address complicated problems,” Aghasadeghi said, “It is simply a matter of taking the initiative to learn about other fields and talk to other professionals. That is the direction that a lot of research is going today, and I think it is very exciting.”

Daken is a junior in ACES. He can be reached at [email protected].

Editor’s note: A previous version of this article incorrectly stated that prosthetist Kevin Carroll developed the prosthetics used by runner Oscar Pistorius. It should have stated Carroll has helped to develop prosthetics like those used by Pistorius but not those exact ones. The Daily Illini regrets the error.