Vanderbilt University Develops First Powered Knee and Ankle Prosthesis

The Vanderbilt prosthesis, the first powered knee and ankle
transfemoral device, is a 7-year effort directed by Michael Goldfarb, H. Fort
Flowers professor of mechanical engineering, Vanderbilt University. The
Vanderbilt prosthesis distinguishes itself from traditional microprocessor
prostheses by requiring less energy from the user to walk. Designed for daily
life, the device adapts and adjusts its biomechanics based on terrain and user

“Users can more easily and without significant effort do a stair
over stair ascent because it is contributing power,” Goldfarb told
O&P Business News. “Amputees with passive microprocessor
controlled knees could go up stairs but they are not getting any assistance
from their device. The Vanderbilt prosthesis provides full assistance.”

Amputees can descend down steps in a passive microprocessor prosthesis,
according to Goldfarb. The problem is the instability of the ankle.

“That is a significant issue,” Goldfarb said. “If the
amputee tries walking down stairs with a stiff or rigid ankle, the user will
not be stable because the ankle needs to dorsiflex significantly in order to
lower the amputee to the next stair.”

According to Goldfarb, the Vanderbilt prosthesis provides better
biomechanics in a variety of terrain scenarios and it enables the user to walk
with less energy consumption. According to Goldfarb, these are two major

Users of the Vanderbilt prosthesis may feel another distinction. The
Vanderbilt prosthesis weighs approximately two pounds more than the
microprocessor prosthesis. This can be considered a considerable amount of
weight for amputees.

“The Vanderbilt prosthesis weighs more than the typical prosthetic
leg,” he said. “This can be seen by most as a disadvantage. Even
though it weighs more, the Vanderbilt prosthesis powers itself. We need more
studies to ascertain how the added power trades with the added weight. A
motorcycle weighs more than a bicycle, but the motorcycle moves itself and a
bike does not. It is easier for a motorcycle to go up a steep hill even though
a motorcycle weighs more. That is the comparison that you have here. We believe
the net effect is a beneficial one. I say that with the caveat that it has not
been quantified at this point.

How does the user control the power? That is the question Goldfarb and
his research team continually worked and improved upon during its seven
versions. The electronics board of the prosthesis alone has been redone 15
times. Most of the iterations came from the software. According to Goldfarb,
iteration is a fundamental part of engineering design.

“We did not know how many sensors we would need in order to infer
the user’s intent,” he said.

Goldfarb and his team added numerous sensors on the prosthesis believing
it was easier to pare them down rather than not having enough information from
the user.

“Once we had all the information we went through and analyzed what
pieces of data were important,” he said. “We stripped it down to the
essential information. We now have fewer sensors with better components. The
goal was to strip it down, make it simpler and improve its overall

Because the device can move on its own, the communication between the
user and the Vanderbilt prosthesis must be reliable and seamless. The device
must be implicit and infer its actions based on how the user is moving. Another
complication for Goldfarb and his team is creating a system that can control
both the knee and ankle joints. No one else in the field has developed a
powered knee and ankle. Can the sensors safely control two joints based on the
movement of the user?

“The device has accelerometers, gyroscopes, load sensors and it
knows its own configurations,” Goldfarb said. “Based on those
sensors, it looks for patterns, and adjusts its behavior based on patterns of
how the person is moving. This has proved to be reliable.”

Although the device needs more study, Goldfarb hypothesizes that the
Vanderbilt prosthesis will substantially decrease the likelihood of amputee
falls. He incorporated numerous behaviors that will help protect the user from
falling. Although the device is considered in the final stages of development,
Goldfarb fully expects to study the device further and produce more substantial

“We need more study on this specific subject, but I would
hypothesize that this leg would relieve somebody’s dependence on
concentration of each step,” he said. “The concentration and
cognition that is required to walk would be handled by the leg.” —
by Anthony Calabro

Disclosure: Goldfarb is a professor at
Vanderbilt University.


From an education point of view, we will discuss this technology, but it
is not something we would discuss at length with the students. From a
university perspective, when we instruct the students, it must be regarding
information that has been researched, with quite a bit of evidence-based
practice to it. In our field, that is lacking anyway.

We still like to promote the newer technologies so the students are
aware of what they may see in their residencies and beyond. For instance, we
will talk about microprocessor knees. It would be nice for the students to fit
those so they have some experience with those devices, although most
universities can not afford it.

  Thomas Karolewski
  Thomas Karolewski

I completely see thought-controlled technology coming down the way in
the future. I can easily see a myo-controlled microprocessor knee perhaps
coupled with an i-Walk powered ankle. We are headed in that direction as we
move forward and there are some really spectacular designs coming out from all
parts of the world.

From a school’s point of view, we still have to be careful.
Promoting cutting-edge technology is great and the students should be aware of
it. But we also have to make sure the students are getting the minimum
standards at the very least so they can pass their exams.

— Thomas Karolewski, CP,

O&P Business News Practitioner Advisory Board
Member Director of Prosthetics Northwestern University Prosthetics-Orthotics

Leave a Reply

Your email address will not be published.