AI-guided brain stimulation helps memory in traumatic brain injury

Now a new study in the journal brain stimulation shows that targeted electrical stimulation in patients with head trauma resulted in an average 19% increase in word recall.

Led by University of Pennsylvania psychology professor Michael Jacob Kahana, a team of neuroscientists studied TBI patients with implanted electrodes, analyzed neural data as the patients studied words, and used a machine learning algorithm to predict momentary memory loss. Other lead authors include Wesleyan University psychology professor Youssef Ezzyat and Penn research scientist Paul Wanda.

“The past decade has seen tremendous advances in the use of brain stimulation as therapy for several neurological and psychiatric disorders, including epilepsy, Parkinson’s disease and depression,” Kahana said. “Memory loss, however, is an enormous burden on society. We lack effective therapies for the 27 million Americans who suffer.”

Study co-author Ramon Diaz-Arrastia, director of Penn Medicine’s Traumatic Brain Injury Clinical Research Center, says the technology developed by Kahana and his team provides “the right stimulation at the right time, informed by the individual’s brain wiring and that the individual’s successful memory retrieval.”

He says the main causes of TBI are road accidents, which are decreasing, and falls, which are increasing due to an aging population. This is followed by aggression and head trauma related to participation in contact sports.

This new study builds on previous work by Ezzyat, Kahana and their collaborators. Publishing their findings in 2017, they showed that stimulation delivered when memory is thought to be failing can improve memory, while stimulation delivered during periods of good functioning worsens memory. The stimulation in this study was open-loop, meaning it was applied by a computer without regard to the state of the brain.

In a study of 25 epileptic patients and published the following year, they monitored brain activity in real time and used closed-loop stimulation, applying electrical impulses to the left lateral temporal cortex only when memory was thought to fail. They found a 15% improvement in the likelihood of remembering a word from a list.

But the new study focuses specifically on eight people with a history of moderate to severe CBT, who were recruited from a larger group of patients undergoing neurosurgical evaluation for epilepsy. Seven of the eight are men, and Diaz-Arrastia says 80% of people hospitalized with traumatic brain injury are men.

Kahana stresses the importance of treating TBI-related memory loss, noting that “these patients are often relatively young and physically healthy, but they face decades of impaired memory and executive function.”

The researchers’ main question was whether stimulation could improve memory across entire lists of words when only certain words were stimulated, whereas previous studies only considered the effect of stimulation on individual words. Ezzyat says this development is important because “it suggests that possible real-life therapy could provide more generalized memory enhancement — not just at the precise moment stimulation is triggered.”

The study notes that more work needs to be done before this type of stimulation can be applied in a therapeutic setting, and scientists need to study the physiological responses to stimulation to better understand the neural mechanisms behind the stimulation. improved memory performance. Diaz-Arrastia says, “It’s still the early days in the field.”

“I think ultimately what we would need,” he says, “is a self-contained implantable system, where you could implant the electrodes into the brain of somebody who’s had a brain injury.”

Michael Kahana is the Edmund J. and Louise W. Kahn Professor of Psychology at the University of Pennsylvania.

Ramon Diaz-Arrastia is Director of the Traumatic Brain Injury Clinical Research Center, Associate Director of Clinical Research at the Center for Neurodegeneration and Repair, and John McCrea Dickson MD Presidential Professor at Penn’s Perelman School of Medicine.

Youssef Ezzyat is an assistant professor of psychology at Wesleyan University and a former senior data scientist in Kahana’s lab at Penn.

Other co-authors include Penn’s Paul A. Wanda, Ethan A. Solomon, and Richard Adamovich-Zeitlin; Bradley C. Lega and Kan Ding of the University of Texas Southwestern; Barbara C. Jobst of Dartmouth-Hitchcock Medical Center; and Robert E. Gross of Emory University.

This research was supported by the Defense Advanced Research Projects Agency (N66001-14-2-4032).