Using intentions to control a robotic prosthetic

A neural implant on the area of the brain that controls the patient's intention to move could be the key to better robotic prosthetics.

Michelle Starr Science editor
Michelle Starr is CNET's science editor, and she hopes to get you as enthralled with the wonders of the universe as she is. When she's not daydreaming about flying through space, she's daydreaming about bats.
Michelle Starr
4 min read

Patient Erik G. Sorto enjoying a beer, unassisted. Caltech, Keck USC, Rancho Los Amigos Team

Researchers around the globe have been attempting to create a mind-controlled prosthetic which would allow people with paralysis or amputations to control the movements of a robotic limb. One method of achieving this involves a neural implant attached to the part of the brain that controls movement, but this often results in jerky, lagged motion in the robotic limb.

But a neural implant attached to a different part of the brain has resulted in a smoother, more natural control over the prosthetic. A team of scientists from the California Institute of Technology on Thursday announced a successful trial with a prosthetic that attaches to the posterior parietal cortex (PPC), or the part of the brain that controls not movement itself, but the patient's intention to move.

"When you move your arm, you really don't think about which muscles to activate and the details of the movement -- such as lift the arm, extend the arm, grasp the cup, close the hand around the cup, and so on. Instead, you think about the goal of the movement. For example, 'I want to pick up that cup of water,'" said principal investigator Richard Anderson, James G. Boswell Professor of Neuroscience at the California Institute of Technology.

"So in this trial, we were successfully able to decode these actual intents, by asking the subject to simply imagine the movement as a whole, rather than breaking it down into myriad components."

The subject in question, Erik G. Sorto, was the first to participate in a clinical trial of the implant. A paralysed quadriplegic for over 10 years, after implantation, Sorto was successfully able to control a robotic limb to bring a drink to his mouth, shake hands and play Rock, Paper, Scissors -- all with smooth, fluid movements.

A physical motion makes its way through several regions of the brain. The intention to move is first formed in the PPC, a high-level cognitive area of the brain. This intention is then transmitted to the motor cortex, which distributes the signals to the necessary muscles, which move to grasp and pick up the object.

The motor cortex is often the last stop for neuronal signals for patients paralysed by high spinal cord injuries. It seemed a logical location for an implant, which sends the specific signals to a computer to be decoded and sent to a prosthetic. However, the signals at the motor cortex are highly complex, and decoding them results in jerky movements in the prosthetics.

Clockwise from top left: Professor Richard Anderson; Dr Charles Y. Liu, professor of neurological surgery, neurology, and biomedical engineering at USC; patient Erik G. Sorto; and Dr Mindy Aisen, chief medical officer at Rancho Los Amigos, who led the study's rehabilitation team. Lance Hayashida/Caltech

Professor Anderson and his team wanted to see if the simpler signals at the PPC could result in a more natural prosthetic.

"The PPC is earlier in the pathway, so signals there are more related to movement planning -- what you actually intend to do -- rather than the details of the movement execution. We hoped that the signals from the PPC would be easier for the patients to use, ultimately making the movement process more intuitive," Anderson said.

For the clinical trial, the Caltech team worked with surgeons at the University of Southern California's Keck School of Medicine to implant two small electrode arrays in the patient's PPC. Each array consisted of 96 active electrons. Each of these electrons records the activity of a single neuron in the PPC.

Patient Erik G. Sorto, using the robotic arm to give himself a drink. Spencer Kellis and Christian Klaes/Caltech

The arrays were then connected via cables to computers that could process and decode the neural signals recorded by the electrons. These decoded signals were then sent to output -- a cursor and a robotic arm developed by researchers at Johns Hopkins University.

Following recovery from the surgery, Sorto was trained to use the cursor and the arm.

"It was a big surprise that the patient was able to control the limb on day one -- the very first day he tried. This attests to how intuitive the control is when using PPC activity," Andersen said.

Sorto refined his control over the arm, which in turn allowed the team to assess what works and what doesn't. For example, breaking down the movement with a thought such as "I need to slowly move my hand towards the object" won't work, whereas "I want to grab that cup" does, resulting in a much simpler, more intuitive approach -- which is exactly what the researchers hoped using an implant on the PPC.

The team will continue to work with Sorto to refine the technique. Already, however, Sorto said, it has made a big difference to his life.

"This study has been very meaningful to me. As much as the project needed me, I needed the project. The project has made a huge difference in my life. It gives me great pleasure to be part of the solution for improving paralysed patients' lives," he said.

Future research will involve finding solutions for these more practical activities. Currently, Sorto uses his eyes to be able control the arm, but if the technology could also implement touch, it would make fine motor movements much easier to execute. The team also hopes to be able to enable more complex tasks.

"Our future studies will investigate ways to combine the detailed motor cortex signals with more cognitive PPC signals to take advantage of each area's specialisations," Anderson said.