The system uses machine learning to control the computer interface with the person’s brain. Unlike most brain-computer interfaces (BCIs), AI worked without the need for extensive daily retraining.
Senior author Karunesh Ganguly, an associate professor in the UC San Francisco Department of Neurology, described the success in a statement:
The BCI field has made a lot of progress in recent years, but because existing systems have had to reset and recalculate each day, they have not been able to tap into the brain’s natural learning processes. It is asking someone to learn to ride a bike repeatedly from scratch. Adopting an artificial learning system to work smoothly with the brain’s sophisticated long-term learning schema is something that has never been shown before in a paralyzed person.
[Read: We asked 3 CEOs what tech trends will dominate post-COVID]
The system uses an electrocortography (ECoG) array about the size of a Post-It note. The array is placed directly on the surface of the brain, where it monitors electrical activity from the cerebral cortex.
Researchers claim that the system provides long-term, stable recordings of neural activity. This gives it an advantage BCI involves the fast tissues that enter the brain tissue, as they tend to change or lose signal over time.
The team tested the system on a paralyzed person of all four limbs, who used it to control a computer cursor on a screen. First, he asked the user to visualize the movements of his neck and wrist while observing the movement of the cursor. This led the algorithm to gradually update itself so that it could match brain activity with cursor movements.
However, this time-consuming process restricted user control. The researchers therefore tried a different approach: allowing the algorithm to continue to update without a daily reset.
Ganguly said that this has led to continuous improvement in system performance:
We found that we could further improve learning by making sure that the algorithm was not updating faster than the brain – a rate of once every 10 seconds. We see this as an attempt to create a partnership between two learning systems – the brain and the computer – that eventually lets the artificial interface become an extension of the user, such as their own hands or hands.
As the test progressed, the user’s brain began to amplify the pattern of neural activity that moved the cursor. Eventually, he developed a psychic mental “model” to control the interface. Researchers turned off the update of the algorithm, so the participant could use the system without the need for daily adjustments.
When the system maintained its performance for 44 days without any performance or daily practice, researchers began adding additional capabilities for BCI – such as “clicking” a virtual button – without dipping the display.
Ganguly now hopes to use ECOG recordings in more complex robotic systems, including prostheses.
“We always think of the need to design technology that doesn’t end in a drawer, so to speak, but which will actually improve the daily lives of paralyzed patients,” he said. “These data suggest that ECOG-based BCI may be the foundation of such technology.”
you can read Research Paper In the magazine Nature Biotechnology.
So are you interested in AI? Again Join our online program, TNW2020, Where you will hear how artificial intelligence is transforming industries and businesses.
Published September 7, 2020 – 17:20 UTC