Paralyzed man moves robot arm with his mind

UC San Francisco researchers paralyze a man to control the robotic arm through devices that deliver his brain signals to a computer.

He can master, move and drop objects just by imagining himself performing actions.

The device, known as the Brain Computer Interface (BCI), worked for 7 months in 7 months without adjustment. So far, such equipment has only worked for one or two days.

https://www.youtube.com/watch?v=5xivizsi44g

BCI relies on an AI model that can adapt to small changes in the brain’s repeated movements in a person – or in this case an imagined movement – and learn to move in a more refined way.

“The integration of learning between humans and artificial intelligence is the next stage,” said Karunesh Ganguly, professor of neurology and member of the Institute of Neuroscience at UCSF WEILL, Karunesh Ganguly, MD, Karunesh Ganguly, MD, Dr. Karunesh Ganguly, MD, Dr. Karunesh Ganguly, MD, said. “This is what we need to achieve complex, lifelike functions.”

The study, funded by the National Institutes of Health, appeared in a cell on March 6.

The key is that the study participants repeatedly imagine how activities change in the brain when they make specific movements. Once the AI ​​is programmed to explain these transitions, it works for several months at a time.

Position, position, position

Ganguly studied how patterns of brain activity in animals represent specific movements and found that the daily changes in these representations as animals learn. He suspects that the same thing is happening to humans, which is why their BCIs lose their ability to recognize these patterns so quickly.

Ganguly and Neurology Researcher Dr. Nikhilesh Natraj collaborated with a study participant who had been paralyzed from stroke for many years. He cannot speak or move.

Tiny sensors were implanted on the surface of his brain that he might absorb brain activity when he imagined moving.

To see if his brain pattern changes over time, Gangali asked the participants to imagine different parts of their body, such as his hands, feet or head.

Although he can’t actually move, the participant’s brain can still generate signals of movement when he imagines himself doing so. BCI recorded how the brain performed on these movements through his brain sensors.

Ganguly’s team found that the shape of the manifestations in the brain remained the same, but their position changed slightly every day.

From virtual to reality

Ganguly then asked the participants to imagine themselves performing simple movements with their fingers, hands or thumbs over a two-week period, while the sensors recorded his brain activity to train the AI.

The participants then tried to control the robot’s arms and hands. But the movement is still not very accurate.

Therefore, Ganguly practiced the participant’s exercises on the virtual robot arm, which gave him feedback on the accuracy of his visualization. Eventually, he got the virtual arm to do what he wanted to do.

Once the participant started practicing with a real robot arm, it only took him some practice to transfer his skills to the real world.

He can pick up the robot’s arms and move them to a new position. He was even able to open a cupboard, take out the cup and secure it to the water dispenser.

Several months later, the participant was still able to control the robot arm after 15 minutes of “adjustment” to adjust how the motion representation drifted since he began using the device.

Ganguly is now perfecting the AI ​​model to make robotic arms move faster and smoother, and plans to test BCI in a home environment.

For paralyzed people, the ability to feed or drink water will change lives.

Ganguly thinks this is reachable.

“I’m very confident that we have now learned how to build systems and that we can do the work,” he said.

Author: Other authors of this study include Sarah Seko and Adelyn Tu-chan of UCSF and Reza Abiri of the University of Rhode Island.

Funding: This work was supported by the National Institutes of Health (1 DP2 HD087955) and the UCSF WEILL Institute of Neuroscience.