A groundbreaking new brain implant has allowed a man with ALS to speak and sing using his ‘real voice’ for the first time. The brain-computer interface (BCI) translates his neural activity into words almost instantly, enabling him to convey changes in tone, emphasize words, and hum melodies in three pitches.
This innovative system, described in a recent study published in Nature magazine, uses artificial intelligence to decode the participant’s electrical brain activity as he attempts to speak. Unlike previous BCIs that had a delay in speech production, this device reproduces not only the intended words but also features of natural speech such as tone, pitch, and emphasis, which are essential for expressing meaning and emotion.
The study participant, a 45-year-old man with ALS, underwent surgery to implant 256 silicon electrodes in a brain region that controls movement. These electrodes captured his brain signals every 10 milliseconds, allowing the BCI to decode the sounds he attempted to produce in real time. The system personalized the synthetic voice to sound like the man’s own by training AI algorithms on recordings of his voice before the onset of his disease.
Using the device, the participant was able to spell out words, respond to open-ended questions, and express himself freely, even using words that were not part of the decoder’s training data. He reported feeling happy and that it felt like his ‘real voice’ when listening to the synthetic voice produce his speech.
The BCI also identified whether the participant was asking a question or making a statement, and could adjust the tone of his synthetic voice based on the emphasis he placed on different words in a sentence. This level of sophistication in speech production represents a significant advancement in BCI technology and opens up possibilities for daily use by patients in the future.
Overall, this new brain implant has the potential to revolutionize communication for individuals with severe speech disabilities, offering them the freedom to express themselves in their own ‘real voice’. This remarkable achievement marks a major milestone in speech BCIs and paves the way for further advancements in assistive technology for those in need.
