Paralyzed People Can Control Robotic Arms With Power Of Mind And Improved Brain-Machine Interfaces, Says Study
Any learning, even playing the piano, will begin as a struggle where hitting the wrong keys is a very common mistake. But improvement comes over as training and practice progress with jerky movements becoming sharp and smooth.
Such an improvement in learning makes a start at the brain and works in tandem with neurons in delivering greater motor control through better management of neural circuits.
Neural Activity Patterns
A new study by neuroscientists from centers such as the Champalimaud Centre for the Unknown in Lisbon and the University of California-Berkeley examined the patterns of neural activity ingrained in the motor learning process of animals.
The findings have been reported in Neuron.
The research results have encouraged the scientists in suggesting that brain-machine interfaces (BMI) can be optimized so that paralyzed patients can use robotic arms with the power of the mind.
The researchers studied the evolution of the motor learning process from the start as a trial and error process until it consolidates in fair proficiency.
"When learning a novel motor skill, animals initially have to explore different movements, " said Vivek Athalye, the lead author.
The research team extrapolated the study to examine the process by which the brain drives the activity of neurons in bringing about concrete behavioral changes.
Experiments And Results
For a data-based analysis, the team roped in neuroscientist Rui Costa and made use of the data compiled by Karunesh Ganguly and José Carmena who co-authored the study.
Experiments were conducted on animals with implanted electrodes in the motor cortex aligned with a BMI that recorded the activity of motor cortex neurons while the animals learned to move a cursor.
From the trial phase to consolidation, the learning progressed until the animals made proficient cursor movements after a series of training sessions.
The changes in neural activity at the motor cortex were monitored during other tasks as well.
They drew the conclusion that there is high variation in motor cortex activity at the start of learning though it turns consistent later on.
From those observations, the researchers turned the gaze on how the brain handles this mechanism. Clarity was lacking whether neurons varied the activity independently or it was a collective change led by many neurons.
That created a new focus on how neurons coordinate learning. The puzzle was how activity patterns were acquired by neurons and in pinpointing that as an autonomous operation or coordinated activity.
Extrapolating the trends from the results, Carmena said it was evident that better BMIs can do well in medical applications such as using robotic arms by paralyzed patients.
Theoretically, this can be done by filtering out the crowd of neurons that are not necessitated in an action and prioritizing those neurons relevant to the performance of the motor task. This is how the robotic arm will work by sending them the relevant signals.
The difference with natural motor learning is that the muscles make the movements under the direction of the brain. Also, it is unknown which neurons control the muscles. However, in the experiments with animals, only those neurons who really do the job were connected to the BMI.
The authors are positive that a similar pattern can work in natural motor learning and expand the applications in many areas.
Robots In Operating Rooms
Meanwhile, operating rooms will soon see more robots in the surgical team for precise and automated surgeries.
Michael Yip, electrical engineering professor at the University of California San Diego, has set up a lab where engineers are developing advanced robotic systems to that end.
It is based on the premise that intelligent algorithms will enable robots to support surgery using smart endoscopes that maneuver through intricate interiors of the body for improving the capabilities of surgeons.