Thinking About Moving

by Guy McKhann, M.D.

May 30, 2013

Over the years we have learned the steps involved as our brain controls our motor system. Let’s say you want to reach for and drink a cup of coffee. First you have to see it and recognize it before you can give a command to reach for it. If you want to use your right hand, you’d activate the neurons in the left motor cortex involved with reaching. While doing that, you shape your hand to conform to the cup’s handle. You pinch your fingers around the handle, grasp it, and bring the cup to your lips. That’s all on the motor side. On the sensory side, there is the feedback from vision and touch as to where your hand is, and importantly, the sense of touch from the handle of the cup. If the cup is too hot, you pull your hand away. Lastly you sense when the cup is touching your lips.

There are a lot of reasons why this system might go wrong. If you’re blind, you must rely on touch to find the cup. But what if the signals sent from the neurons that control the movements of your right hand don’t travel down the spinal cord properly? What if your hand is missing? All of these happen in such disease processes as stroke, spinal cord injury, peripheral nerve injury, and traumatic amputation.

Much of physical therapy is directed at getting the existing system to recover at least part of its previous functions. And in many cases physical therapy is helpful. But in some instances, such as spinal cord injury, severe nerve injury, or amputation, the existing system is either disconnected or gone, never to return.

What can be done? Assuming that the neurons in the left motor cortex are functioning, maybe you can encode these neuronal signals and bypass the rest of the outflow system, transmitting the neuronal signals to a robotic limb. In other words, get a robotic limb to move just by thinking about it. This is the outcome described in the article “Mind-controlled Prostheses Offer Hope for Disabled.”

To achieve this, a recording device is placed on or in the brain. Tiny electrodes can record signals from the designated motor neurons, and these signals can be picked up by the chip and transmitted outside the brain to computers which can carry out functions like moving a cursor or sending signals to a robotic limb.

What was on the forefront 3-5 years ago is now being accomplished in a number of laboratories across the country. As the field is shaking out a little, it is appropriate to make some comments on what might be coming next. I am aided in this by a superb review article by Jose Carmena (“How to control a prosthesis with your mind”) in IEEE Spectrum.

Originally the focus was on recording from neurons in the motor cortex. This may not be necessary. As we learn a motor movement, over time and after repeated use, we develop a motor circuit for that movement. Athletes do this all the time. A tennis player does not think about the details of his serve, he just hits it where he is aiming. He has developed a motor program for the serving, commonly referred to as muscle memory. The same happens with a pianist. A colleague of mine, a concert-level pianist, told me that when learning a new work, initially he played a part of a piece very slowly, over and over again, gradually speeding up. Finally he played up to speed without even thinking about what he was playing. He had developed a motor program which probably was no longer based only in the motor cortex, but also involved deeper structures in the brain, such as the basal ganglia. Thus it should be possible for a prosthetic system to tap into a motor program at several sites in the brain. Further, once the prosthetic system learns where to go, it would be as if it had memorized what at to do, not having to relearn with each repeated use.

In describing what we normally do with a motor movement, I emphasized that it is a two-way street. There is the outflow of motor commands, but also the input of sensory system that gives information about shape, surface properties, speed of movement, and place. Without this feedback, our movements are all over the place, controlled by sight, if possible. A challenge to developing mind-controlled prosthetic systems is to incorporate this sensory feedback.

One can try to describe what is going on in this field, but the videos of what subjects are starting to do are spectacular. Go on the web under “Mind Controlled Prostheses” and you will find many examples.