Maryam Shanechi, the Sawchuk Chair in Electrical and Computer Engineering and founding director of the USC Center for Neurotechnology, and her team have developed a brand new AI algorithm that may separate brain patterns related to a selected behavior. This work, which may improve brain-computer interfaces and discover latest brain patterns, has been published within the journal Nature Neuroscience.
As you’re reading this story, your brain is involved in multiple behaviors.
Perhaps you’re moving your arm to grab a cup of coffee, while reading the article out loud in your colleague, and feeling a bit hungry. All these different behaviors, corresponding to arm movements, speech and different internal states corresponding to hunger, are concurrently encoded in your brain. This simultaneous encoding gives rise to very complex and mixed-up patterns within the brain’s electrical activity. Thus, a serious challenge is to dissociate those brain patterns that encode a selected behavior, corresponding to arm movement, from all other brain patterns.
For instance, this dissociation is essential for developing brain-computer interfaces that aim to revive movement in paralyzed patients. When serious about making a movement, these patients cannot communicate their thoughts to their muscles. To revive function in these patients, brain-computer interfaces decode the planned movement directly from their brain activity and translate that to moving an external device, corresponding to a robotic arm or computer cursor.
Shanechi and her former Ph.D. student, Omid Sani, who’s now a research associate in her lab, developed a brand new AI algorithm that addresses this challenge. The algorithm is known as DPAD, for “Dissociative Prioritized Evaluation of Dynamics.”
“Our AI algorithm, named DPAD, dissociates those brain patterns that encode a selected behavior of interest corresponding to arm movement from all the opposite brain patterns which are happening at the identical time,” Shanechi said. “This permits us to decode movements from brain activity more accurately than prior methods, which may enhance brain-computer interfaces. Further, our method may also discover latest patterns within the brain that will otherwise be missed.”
“A key element within the AI algorithm is to first search for brain patterns which are related to the behavior of interest and learn these patterns with priority during training of a deep neural network,” Sani added. “After doing so, the algorithm can later learn all remaining patterns in order that they don’t mask or confound the behavior-related patterns. Furthermore, the usage of neural networks gives ample flexibility when it comes to the sorts of brain patterns that the algorithm can describe.”
Along with movement, this algorithm has the pliability to potentially be utilized in the longer term to decode mental states corresponding to pain or depressed mood. Doing so may help higher treat mental health conditions by tracking a patient’s symptom states as feedback to exactly tailor their therapies to their needs.
“We’re very excited to develop and show extensions of our method that may track symptom states in mental health conditions,” Shanechi said. “Doing so could lead on to brain-computer interfaces not just for movement disorders and paralysis, but in addition for mental health conditions.”
