Brain-Computer Interface Lets Users Learn to Move Cursor in Seconds

Brain-computer interfaces (BCIs) allow severely disabled people to control wheelchairs, robotic arms, and of course computers. While much progress has been achieved toward improving the accuracy and precision of these devices, they have required long periods of tedious training for users to get acquainted with the technology. The computer has to be taught to understand each user’s unique electrical activity patterns that code for desired movement that the person wants to perform.

Now a team of researchers has come up with something incredible and unexpected. They’ve come up with a way of “calibrating” a BCI so that a new user is able to start reliably using it in a matter of seconds. As an example, in a newly published study in Journal of Neural Engineering, a 63-year-old man who has never used a BCI before was able to move a cursor on a screen toward lit-up targets less than a minute after calibration.

The BCI employed was the BrainGate device which was used in previous achievements. Before, it had to involve special technicians and training sessions to achieve the level of control that is now possible after a short, self-regulated calibration process. The calibration can even be seen, as users begin to achieve control of a cursor within seconds of initiation, quickly improving for at least the first three minutes of the process.

As the users imagine moving a cursor using their hand, the calibration software uses statistical learning methods to quickly understand what the neural signals arising from the motor cortex are intended for. Repeating this process by having the user wish to move the cursor to different parts of the screen allows the software to tune into the person’s wishes.

The technology will certainly help advance this new and exciting field, and will hopefully help a lot of severely paralyzed people gain a new level of independence.

Study in Journal of Neural Engineering: Rapid calibration of an intracortical brain–computer interface for people with tetraplegia…

Link: BrainGate project page…

Via: Brown University…