Paralyzed People Successfully Navigate Rooms Using Mind-Controlled Wheelchair
By Mikelle Leow, 21 Nov 2022
Paralyzed patients can move around on their own volition with a wheelchair that’s powered by their thoughts, coupled with training.
Explaining in the iScience scientific journal, José Millán at the University of Texas in Austin and his team say they have taught three people with severe or full paralysis in their limbs to get around a room with a brain-computer interface plugged into a wheelchair, while avoiding obstacles, to a noteworthy level of success.
The experiment is a follow-up to two others involving non-disabled persons. One of them had the subject fixate on flickering lights in various points of a room, which were converted into brain signals to tell artificial intelligence which direction to steer the wheelchair towards. The other saw electrodes implanted into the brains of abled people.
The above methods were either uncomfortable for the eyes—try staring at a bright light for just a few seconds—or required an invasive procedure. In the new test, the mind-controlled wheelchair merely relied on participants’ cognitive abilities to steer them from point to point.
The electric wheelchair is built to travel forward by default. If the user envisions moving both legs, however, it would steer to the left. If they imagine moving both arms, it would turn right.
Subjects wore skullcaps consisting of 31 electrodes to send those signals non-invasively. For more accuracy, the headwear was adhered to their heads with a gel.
The thought of moving one’s limbs was transmitted to a laptop strapped to the back of the wheelchair, to be interpreted by an AI and translated into a wheel movement.
The trio were trained to refine their signals three times a week over a course of two to five months, and each emerged with varying results.
‘Person 1’ improved their accuracy from 37% to 87% by their last 10 sessions, and ‘Person 3’ managed to pick up their performance from 67% to 91% after training.
‘Person 2’, on the other hand, consistently maneuvered around the room at an average precision of 68%.
In a 15-meter (49-foot) hospital room filled with beds, chairs, and medical equipment, ‘Person 1’ passed through the final checkpoint in an average of four minutes throughout 29 tries, ‘Person 3’ completed their tests on an average of seven minutes throughout 11 attempts, and ‘Person 2’ got through the third checkpoint in an average of five minutes but did not complete the full course.
Millán concludes that the learning curve to make brain signals clearer will be different across the board, but ultimately, paralyzed users will be able to learn to move independently on electric wheelchairs.
“We show that mutual learning of both the user and the brain-machine interface algorithm are both important for users to successfully operate such wheelchairs,” Millán adds.
It’s worth noting that, although the mind-controlled wheelchair may work in cluttered rooms, it might struggle with outdoor environments.
Further, the gel that helps the skullcap stick to the head dries out within a few hours, so users will only be able to control their wheelchair via their mind for so long.
With that being said, paralyzed patients will reclaim some of their power to navigate places on their own, which is a win.
The study follows a recent breakthrough in which fully paralyzed patients have been able to walk again after physiotherapy and the right neurons were activated.
[via New Scientist and SciTechDaily, cover image 154346047 © Beselialuka | Dreamstime.com]