by A groundbreaking study conducted by a team of neurosurgeons and artificial intelligence (AI) specialists at the University of California, San Francisco (UCSF), has shown promising results in helping a nearly mute man regain his speech abilities in two languages.
The team, whose research was published in the journal Nature Biomedical Engineering, implanted a brain-computer interface (BCI) inside the skull of a man, known as Pancho, who had lost his speech capabilities due to a stroke. They then employed AI techniques to interpret the data provided by the BCI to facilitate speech in Spanish and English.
Building upon previous research that demonstrated the feasibility of implanting a probe onto the brain’s surface to read brain waves and convert them into words, the UCSF team took the innovation a step further by introducing a second language into the equation.
Pancho, a native Spanish speaker, had lost most of his Spanish language abilities following a stroke at the age of 20. However, he had learned to read and convert words in his thoughts to English several years later. In this study, the research team applied a lattice of electrodes to a part of Pancho’s brain responsible for language processing. A connector in his skull enabled the BCI to link to a computer system.
Over the course of three years, Pancho underwent extensive training. He was presented with words on a computer screen and was then asked to repeat them silently in his mind. The probe read his brain waves, attempting to convert them into the corresponding word. During the training, Pancho was shown both Spanish and English words. An AI language model (LLM) was employed to aid in deciphering and converting the words, reducing the number of errors.
The system demonstrated an impressive 88% accuracy in distinguishing between Spanish and English, and a 75% accuracy rate in decoding words overall. While these figures may not seem perfect, they were sufficient for Pancho to engage in meaningful conversations with the research team. This breakthrough represents a significant step forward in the field of brain-computer interfaces and language restoration.
