Researchers at UC San Francisco have used an electrocorticography (ECoG) implant to develop a brain-computer interface that does not need to be recalibrated and retrained each time it is used, allowing an experienced user to plug in and begin using the system at any time. The technique could allow for brain controlled prosthetic limbs or wheelchairs for disabled people.
“The brain computer interface field has made great progress in recent years, but because existing systems have had to be reset and recalibrated each day, they haven’t been able to tap into the brain’s natural learning processes. It’s like asking someone to learn to ride a bike over and over again from scratch,” said Karunesh Ganguly, a researcher involved in the study. “Adapting an artificial learning system to work smoothly with the brain’s sophisticated long-term learning schemas is something that’s never been shown before in a person with paralysis.”
A key aspect of the success of the new system is the ECoG array used by the researchers as a brain implant. The notepad sized electrode array is surgically placed on the brain surface, and is less invasive compared with traditional brain implants which penetrate the brain tissue and resemble a pin cushion. These traditional implants tend to move about in the tissue, making the signals they transmit somewhat unstable. They also suffer from immune rejection as the implants cause tissue damage. While the ECoG array is less sensitive, it appears to provide a more stable signal over time, making it more suitable for a plug and play system.
Brain rendering showing weighting of ECoG electrodes that drove BCI control
In this latest study, the researchers tested their system with the help of a volunteer with paralyzed limbs. The volunteer used the system to move a cursor on a computer screen, and a machine learning algorithm helped to match the user’s desired actions with the movements of the cursor. Typically, such an algorithm would need to be reset each time a user wanted to use the system, but the researchers allowed the algorithm and the volunteer to refine their partnership over many sessions. Eventually the system allowed the volunteer to plug in and immediately have full control over the cursor.
“We see this as trying to build a partnership between two learning systems — brain and computer — that ultimately lets the artificial interface become an extension of the user, like their own hand or arm,” said Ganguly. “Once the user has established an enduring memory of the solution for controlling the interface, there’s no need for resetting. The brain just rapidly convergences back to the same solution.”
Study in Nature Biotechnology: Plug-and-play control of a brain–computer interface through neural map stabilization