Researchers in the US have developed a communication method for people with paralysis that uses a computer to convert mental handwriting into words on the screen.
As part of a new study, a paralyzed man used his brain to write on a screen at a speed almost as fast as an able-bodied adult texting on a smartphone, according to the developers at Stanford University.
A computer decodes handwriting movements from brain signals and can enable much faster communication than was previously possible.
“This approach allowed a person with paralysis to compose sentences at a speed almost comparable to that of healthy adults of the same age typing on a smartphone,” said Jaimie Henderson, professor of neurosurgery.
“The goal is to restore the ability to communicate through text.” The researchers linked artificial intelligence (AI) software to a device called a brain-computer interface (BCI) implanted in the man’s brain. According to the study, the software could decode information from the computer to convert the man’s thoughts about handwriting into text on a computer screen.
This approach allowed the man to write more than twice as fast as he could using an earlier method developed by the Stanford researchers, who reported those findings in 2017 in the journal e Life.
Researchers said the new findings could lead to further advances that benefit millions of people worldwide, who have lost use of their upper limbs or their ability to speak due to spinal cord injury, strokes, or amyotrophic lateral sclerosis, also known as Lou Gehrig’s disease.
“Think how much of your day is spent on a computer or communicating with another person,” added co-author Krishna Shenoy. “Restoring the ability of people who have lost their independence to interact with computers and others is extremely important, and that’s what makes these types of projects central.”
An image released by the Howard Hughes Medical Institute of a brain computer interface that converts mental handwriting into text on a screen.
According to the scientists, the participant achieved a writing speed of about 18 words per minute with an accuracy of 94.1 percent.
In comparison, able-bodied people of the same age can pronounce about 23 words per minute on a smartphone.
The researchers instructed him to write sentences as if his hand were not paralyzed, by imagining that he was holding a pen on a piece of lined paper. During this exercise, the BCI used a neural network – a type of machine learning – to translate attempted handwriting movements of neural activity into text in real time.
The participant, named T5, lost virtually all movement under the neck due to a spinal cord injury in 2007. Nine years later, Dr. Henderson placed two brain computer interface chips on the left, each the size of a baby aspirin. of his brain.
Each chip in his brain has 100 electrodes that pick up signals from neurons that fire in the portion of the motor cortex – an area of the brain’s outer surface – that controls hand movement.
The experts then sent those neural signals via wires to a computer, where AI algorithms decode the signals and suspect the T5’s intended hand and finger movement.
“This method is a marked improvement over existing communication BCIs that rely on using the brain to move a cursor to ‘type’ words on a screen,” said lead author Frank Willett.
“Trying to write each letter creates a unique activity pattern in the brain, making it easier for the computer to identify with much greater accuracy and speed what is being written.” Experts say further demonstrations of longevity, safety, and efficacy are required before the method can be used clinically.
The researchers suggest that these methods could be more broadly applied to any sequential behavior that cannot be directly perceived, such as decoding the speech of someone who can no longer speak.