Scientists have developed a brain-computer system that enables people with paralysis to communicate and have real-time conversations.
Researchers connect Ann’s brain implant to a computer that creates speech sounds. (Photo by Noah Berger)
A researcher from the University of California, Berkeley, and San Francisco has developed a brain-computer interface that helps people with severe paralysis speak more naturally in a study published in Nature Neuroscience. The innovation highlights how this breaks problems in speech devices and marks real progress since it allows people who have lost the ability to speak to engage in real-time communication.
Researchers have used AI to create a system that immediately turns brain signals into spoken words. It works like voice assistance such as Alexa or Siri, allowing smooth and natural speech in real-time.
According to study co-lead author Cheol Jun Cho, neuroprosthesis collects brain signals from the motor cortex—the area that controls speech—and uses AI to turn those signals into spoken words. The researchers tested this on a 47-year-old woman, Ann, who hadn’t been able to talk for 18 years.
In the clinical trial, doctors placed electrodes on the surface of Ann’s brain to capture her brain activity as she tried to say sentences shown on a screen. An AI system then turned those brain signals into speech, using a voice that sounded like hers before the injury.
While brain-computer interfaces have been around for years, the innovation in this research is its incredible ability to translate thoughts into speech in nearly real time. Previous systems used to have staggering, sometimes taking 8 seconds to speak a sentence.
This AI technology brings researchers closer to achieving naturalistic speech with BCI devices. It is their most significant achievement, and they are optimistic about advancing at every level. So, let’s hope for a better future with innovation and how it can help improve the lives of patients with ALS or paralysis caused by stroke.
