What was the problem you were looking to address with this research?
One of the major challenges of amputation is that the farther away from the wrist the amputation happens, the more functions are lost. At the same time, there are fewer signals to try to replace them. As you move from below elbow to above elbow, it becomes extremely difficult to try to replace all those distal functions. The only thing we can really predict is flexion extension of the biceps because all the finger muscles are below the elbow.
What was the technique you helped develop to address the problem?
We call it a neuromusculoskeletal prosthesis. We started with the osseointegration technology, the bone-anchored implant, then Dr. Max Ortiz-Catalan, the director of our lab, added implantable sensors that go into the muscles and the nerves. Those components allow us to address many of the limitations of prosthetic limbs. The bone anchoring allows for a better range of motion. The implantable sensors allow us to read electrical signals from the muscles whenever they contract, which we can use to predict what someone’s intent is. When someone has an amputation, the muscles are still there; they're just no longer connected to the joints that they used to act upon. But those muscles can still contract. Typically, someone with an amputation above the elbow will flex their bicep or their triceps, we read those electrical signals, and we use that to determine whether someone is, for example, trying to open or close a hand.
How were you able to use the remaining nerves in the upper part of the arm to help the patient gain better control of the bionic hand?
The nerves that used to go down to the finger muscles are still up in the upper arm, but those nerves are no longer connected to anything. Dr. Paolo Sassu, our team’s hand surgeon, split the nerve up into multiple fibers, and in some instances, he would grow that fiber directly into a muscle that was still there but not activated on anything. This is a technique we call targeted muscle reinnervation. The crux of this surgery is that nerves are dumb. You can plug them into any muscle, and whenever the nerve receives a brain signal to do some function, that muscle is going to contract.
In this case, we took some functionality that was related to the pronation and supination of the wrist, or maybe the thumb or a couple of fingers. When our patients would try to perform those movements, that muscle would contract, which meant we can read that electrical signal and use that to control a hand.
How did muscle grafts play a part in the research?
There’s a surgery we use to create what we call a regenerative peripheral nerve interface, or RPNI. It’s a little bit like a bacon-wrapped date. You take a nerve, and you take a muscle graft, either from somewhere else in the body or a donor graft, and you wrap that muscle around the end of the nerve. Whenever that muscle fiber fires, that tiny muscle graft will contract. If you put an electrode inside of that muscle, you can read the same electrical signals. For somebody with an amputation above the elbow, we can use that, so they are able to control a prosthetic hand and move each of the fingers individually. That normally would not be biologically possible, because all those muscles exist below the elbow. But through the surgical reconstruction, we're able to bring those back.
What technology did you use to translate the signals from the muscles into movement in the bionic hand?
After the surgery is done, we ask our participant to try to flex his thumb, his index finger, and his middle finger, and we record all the electrical signals from these muscles and nerve grafts. We take all that data and feed it into a machine-learning algorithm, and the machine-learning algorithm will pick out, “OK, this is what the electrical signals look like when he's trying to flex the index finger. This is what the electrical signals look like when he's trying to extend the index finger.” We teach the computer to recognize these electrical signals, but once it's able to do that, we can have the patient connected to his prosthesis, and whenever he creates these electrical signals, it will handle the movement.
Is the idea that eventually everyone with a new amputation would get this? Or is this something that would be a second step after you've had had the traditional socket prosthetic?
It's hard to say right now. It's still under investigation in the U.S. There are regulations we would need to go through. Also, it's not a perfect fit for everybody. There are weight limitations and bone strength limitations that need to be met.
Here at the CU School of Medicine our approach is that we want people to start with a traditional socket prosthesis, and if they don't have any problems with that, then we leave it alone. But if they're running into problems with that, then we can start discussing whether osseointegration is the right choice for them. As far as the implanted sensors, I think that in a perfect world, every trauma surgeon would know how to do these nerve transfers and how to identify what muscles are most useful and what functions are most useful for a prosthetic hand and can do those at the time of amputation. That way, when the prosthetists and occupational therapists and engineers come along later, the patient is already set up to have the best outcome.
What was the most exciting part of this research for you?
The most exciting part is the functionality this patient has gained from being involved in the study. After receiving the bone-anchored limb and the implanted sensors, he drastically and very quickly improved his functionality and his independence. He participated in a competition called the Cybathlon, which is like the cyborg Olympics. He participated in a couple of races to perform daily tasks, things like tying your shoes and cutting bread and opening jars and hanging up clothes and changing light bulbs. The competition was a huge motivating factor for him.
Do you plan to continue this research at the CU School of Medicine?
Right now, there are no plans to use this particular technology, but I'm interested in leading the lab to doing more of this neural interfacing. I'm no longer formally affiliated with the Swedish lab, so my focus going forward is really on bringing a lot of that expertise in technology and trying to implement that here at CU to turn this campus into a global center for prosthetics, technology, and rehab.