Read the text below.
Researchers from the University of California, San Francisco, created a device that can convert brain signals into speech.
To develop the device, the researchers worked with patients who were about to undergo brain surgery. The device was implanted into the surface of the patients’ brains where it collected brain signals that control the movements of the lips, tongue, jaw, and voice box while speaking. It then transmitted those signals to a computer that transformed brain activity into synthesized speech.
Lead researcher Edward Chang believes that the device serves as a ray of light for people who have lost the ability to speak because of stroke or other conditions.
To check how intelligible the device is, the researchers recruited native English speakers to listen to and transcribe the electronically-generated speech. The device accurately synthesized “sh” and “z” sounds but struggled with “b” and “p.” Nonetheless, the volunteers understood about 70% of the words they heard.
The device can translate 150 words per minute when properly set up. This means that it works much faster than other speech synthesizers, which can produce only up to eight words per minute.
Despite the positive results, the device will not be available for widespread use in the near future. The researchers will still do clinical trials to improve it. For the next step, they are planning to test the device on people who cannot talk at all and whose brains were not trained to speak. However, since implanting the device requires invasive surgery, they might find it challenging to get the right patients.