AI cracks the mind-reading code – what’s next?

The development of brain-computer systems has received a boost after scientists created AI that can read minds.

The human brain constantly manages multiple behaviors every second, controlling our movements, feelings, and internal states. A major challenge for scientists developing brain-computer interfaces has been figuring out how to identify particular behaviour, such as arm or leg movements among all the other electrical activity happening in the brain at the same time.

This is where AI comes in handy. Scientists at the University of Southern California (USC) have now created an algorithm, named DPAD, that can dissociate brain activity patterns that encode a particular behavior of interest.

An AI algorithm learns a particular behavioral pattern during the training of a deep neural network and looks for it thereafter.

"After doing so, the algorithm can later learn all remaining patterns so that they do not mask or confound the behavior-related patterns. Moreover, the use of neural networks gives ample flexibility in terms of the types of brain patterns that the algorithm can describe," explained Omid Sani, a research associate.

According to Maryam Shanechi, lead author and founding director of the USC Center for Neurotechnology, this new approach can “enhance brain-computer interfaces,” and also help to discover new patterns in the brain that may otherwise be missed.

From moving the arm to identifying depression

AI’s ability to read brain activity and identify behaviors related to electrical signals holds tremendous potential for people with physical impairments.

The key for developing brain-computer interfaces that aim to restore movement in paralyzed patients has been identifying the brain pattern that facilitates the movement.

Patients can't directly send their thoughts to their muscles. To help them, brain-computer interfaces must decode their planned movements from brain activity and use that information to control devices like robotic arms or computer cursors.

However, the findings, published in the journal Nature Neuroscience, are not limited only to communication with muscles. The scientists believe that the algorithm that they created has the potential to decode mental states such as pain or depressed mood.