Two paralyzed individuals have successfully typed on a virtual keyboard using an implant that decodes their attempts to move their fingers. One patient was able to type text at a speed 80% faster than that of a healthy person, according to a study published in Nature Neuroscience.

Traditionally, brain-computer interfaces (BCIs) for paralyzed individuals rely on eye tracking or recognition of neural activity related to speech. However, researchers from Mass General Brigham and Brown University proposed that the familiar QWERTY keyboard format would be more user-friendly for many.

Thought Controls the Machine: How AI is Transforming Brain-Computer Interfaces

“The most important thing is to have a variety of options for each patient to tailor the technology to their specific condition and situation,” said study author Justin Jude.

In the study, participants were asked to simulate typing on a QWERTY keyboard. The system reliably read brain impulses, recognizing up to 30 different actions—three for each of the ten fingers.

Two individuals participated in the testing of the BCI device from Blackrock Neurotech:

  • Patient T17 (paralyzed below the neck due to a spinal cord injury) achieved a speed of 47 characters per minute with an accuracy of 81%;
  • Patient T18 (suffering from amyotrophic lateral sclerosis, ALS) reached a speed of 110 characters per minute with an accuracy of 95%.

The stability of the results for the second patient lasted a week, while the first lasted only two days.

Jude noted that the higher performance of one participant could be attributed to the number and placement of electrodes in the brain. T18 had six arrays of contacts implanted in the dorsal (upper) part of the precentral gyrus—about three times more than T17.

In T17, some electrodes were also placed in other areas of the motor cortex to collect speech signals.

Differences in results may also stem from the fact that tetraplegia and ALS affect the brain differently, even though both conditions lead to paralysis.

Jude emphasized that decoding finger movement signals could eventually aid in restoring complex hand movements, including grasping and reaching for objects.

In the future, precise recognition of finger motor skills could help patients regain the ability to control prosthetics for complex tasks like grasping and reaching.

However, the technology must overcome significant regulatory hurdles before it becomes widely available to patients.

It is worth noting that in March, China's regulator approved the country's first neuroimplant for commercial use.