The breakthrough nerve interface reads handwriting from a paralyzed man’s brain

The breakthrough brain interface has allowed men to be paralyzed to “type” at 90 characters per minute, with new research showing an extraordinary leap in communication is possible for those who use the system “brain to text”. Instead of trying to make a virtual keyboard can be used by reading brain nerve activities, the team responsible for the breakthrough focus on tracking imaginary handwriting.

An existing system that tracks brain activity and maps that to a computer it usually relies on thoughts related to arm movements. By tracking it, even if the arm itself cannot move, they can be mapped to highlight keys on the virtual keyboard or other interface types.

When it works, this is limited. The current system allows around 40 characters per minute through the brain computer interface (BCI), according to Krishna Shenoy, the investigation of Howard Hughes medical institutions at Stanford University, with Stanford Neurosurgeon Jaimie Henderson, wrote this new study. Instead of the movement of the arm, they see brain activity when people imagine handwriting.

Apparently, according to HHMI research specialists and Neuroscientist Frank Willett, who worked on the project, imagined the letters to be handwritten in a very distinctive activity pattern. An algorithm trained to recognize them, then, can be much faster than the existing BCI system.

The research subjects were a 65-year-old man who, after a spinal injury, left paralyzed from the neck down. He has two sensors implanted into parts of the brain that usually controls hands and arms. Shortly to the computer, when he imagined writing letters as if by pen on paper, the algorithm could change the impulse into digital text.

“With this system, the man can copy sentences and answer questions with rates similar to someone typed in the smartphone,” the researchers said. In fact, it can produce text at 90 characters per minute, almost double the existing BCI system level.

The aim is to enter mental handwriting as an input option, together than replacing existing point-and-click navigation. The team responsible for new BCI has also worked on greeting decoding, and they imagine an integrated system that will support a number of different input modalities that collectively tapped the superiority of the speed and accuracy of each.

Furthermore, the group plans to work with other participants who cannot speak, because of the capabilities of the system are rounded. Although it is too early for the BCI production version, the aim is to finally allow paralyzed users the ability to communicate in real-time than forcing them to choose more time consuming interfaces.