In the new study, the Stanford team wanted to know whether neurons in the motor cortex contain useful information about speech movements. That is, can they detect how “subject T12” is trying to move her mouth, tongue, and vocal cords when she tries to talk?
These are small, subtle movements, and according to Sabes, a major discovery is that only a few neurons contain enough information to allow a computer program to predict, with great accuracy, the words that The patient is trying to speak. That information is transferred to a computer screen by Shenoy’s team, where the patient’s words appear as they are spoken by the computer.
The new results build on previous work by Edward Chang at the University of California, San Francisco, who wrote a speech concerning the most complex movements that humans make. We push the air out, add audible vibrations, and turn it into speech with our mouths, lips, and tongues. To make the “f” sound, you place your upper teeth on your lower lip and push the air out—just one of the dozens of mouth movements needed to speak.
A road ahead
Chang had previously used electrodes placed on top of the brain to allow a volunteer to talk through a computer, but in their previous preprint, the Stanford researchers say their system is more accurate and fast. three to four times more.
The researchers, including Shenoy and neurosurgeon Jaimie Henderson, write: “Our results suggest a possible avenue to restore communication in people with paralysis at conversational speeds. .
David Moses, who works with Chang’s team at UCSF, said the current work reaches “an impressive new standard of performance”. However, even as records continue to be broken, he said, “it will become increasingly important to demonstrate consistent and reliable performance over many years.” Any commercial brain implant can have a hard time getting past regulators, especially if it degrades over time or if the accuracy of the recording drops.
The way forward is likely to include more complex implants and closer integration with artificial intelligence.