A brain-computer interface that can record and decode an individual's inner monologue has been created by scientists at Stanford University.

The findings may make it easier for those who are nonverbal to interact with others. The novel brain-computer interface eliminates the need for users to try to physically talk, in contrast to certain earlier systems. They simply need to think about what they want to say instead.

Related Brain-to-voice Neuroprosthesis Restores Naturalistic Speech

"This is the first time we've managed to understand what brain activity looks like when you just think about speaking," study co-author Erin Kunz, an electrical engineer at Stanford University, said in a statement. "For people with severe speech and motor impairments, [brain-computer interfaces] capable of decoding inner speech could help them communicate much more easily and more naturally."

A paralyzed person can easily mentally zip their lip and refrain from oversharing by relying on the signals generated when they attempt to speak.  However, it also requires them to exert themselves in order to express a word or sentence, which can be exhausting and time-consuming.

In order to discover a better method, Kunz and a group of researchers examined the brain activity of four individuals who were already communicating with BCIs.

Decoding brain signals that are significantly more subtle than those generated by attempted speech was the team's goal.  The group's goal was to decipher imagined speech.

Even though they are no longer able to, a paralyzed person tries their hardest to physically make intelligible spoken words during attempted speaking.  When someone engages in imagined or inner speech, they simply consider a word or sentence, sometimes by visualizing how it would sound.

The researchers discovered that whereas attempted speech generates stronger signals in the motor cortex, imagined speech produces signals that are identical.  They were also able to convert those weaker signals into words with the aid of artificial intelligence, reports NPR.

"We were able to get up to a 74% accuracy decoding sentences from a 125,000-word vocabulary," Kunz says.

Decoding a person's inner speech made communication faster and easier for the participants. But Kunz says the success raised an uncomfortable question: "If inner speech is similar enough to attempted speech, could it unintentionally leak out when someone is using a BCI?"

According to their research, it might in some situations, such as when someone was silently remembering a series of instructions.

In order to safeguard the privacy of BCI users, the team tested two different approaches.

Initially, they set the gadget to disregard interior speech cues.  That was effective, but it eliminated the ease and speed of decoding interior communication.

According to Kunz, the team adopted a strategy from virtual assistants like Alexa and Siri, which only activate when they hear a particular phrase.

"We picked Chitty Chitty Bang Bang, because it doesn't occur too frequently in conversations and it's highly identifiable," Kunz says.

That allowed participants to control when their inner speech could be decoded.

To restore verbal communication, implants have thus far relied on attempted speech, which necessitates a great deal of user effort.  Additionally, the implants are ineffective for those with locked-in syndrome who are unable to regulate their muscles.  The new decoder taps directly into the brain to capture inner speech, which reduces effort and may facilitate faster communication.