Monkeys transcribe Hamlet with new brain-reading tech

The technology, developed by Stanford University researchers, has been hailed as a major breakthrough for people suffering from severe paralysis such as physicist Stephen Hawking.

According to the team behind the invention, directly reading brain signals to interpret thoughts and drive a computer cursor would allow users to communicate faster than existing technologies allow. For example, the system developed for Hawking by Intel and SwiftKey relies on tracking the movement of facial muscles. Alternatively, eye movement tracking can be used but this doesn’t always work. For example in Hawking’s case, eye movement tracking didn’t work because of droopy eyelids.

"Our results demonstrate that this interface may have great promise for use in people," said Paul Nuyujukian, postdoctoral fellow at Stanford, who developed the system together with professor of electrical engineering Krishna Shenoy. "It enables a typing rate sufficient for a meaningful conversation."

Surprisingly though, the researchers estimate humans will be typing more slowly using the technology than the monkeys involved in the experiment. While the monkeys were just transcribing given passages, humans will be slowed down by thinking about what they actually want to communicate and will also think about how to spell words correctly.

"What we cannot quantify is the cognitive load of figuring out what words you are trying to say," Nuyujukian said.

The technology could be further improved if combined with auto completion technologies, such as those employed in modern smart phones, to improve typing speed.

The experimental monkeys had multi-electrode arrays implanted into their brains already four years ago as the researchers have been testing multiple iterations of the technology over the years. The electrodes, located in the area of the brain that directs hand movements, have thus already proved they can operate over a longer period of time without deteriorating. The researchers said they hadn’t observed any adverse health effects in the animals.

"The interface we tested is exactly what a human would use," Nuyujukian said. "What we had never quantified before was the typing rate that could be achieved."

Over the years, the researchers have been fine-tuning high-performing algorithms responsible for translating the brain signals into the movements of the cursor.

Earlier versions of the technology have already been tested successfully in people with paralysis, but the typing was slow and imprecise.

The results of the study were published in the latest issue of the journal the Proceedings of the IEEE.