Penn’s Restoring Active Memory Project Adds Task and Patient Data to Publicly Available Human Brain Dataset

By Michele W. Berger for PennNews

Penn researchers Michael Kahana and Daniel Rizzuto, who run the Restoring Active Memory project, incorporated data from a new spatial-memory task called “Treasure Hunt” (above). In this video game-like environment, patients navigate around a desert island the size of a football field. When they reach a treasure chest, it opens to reveal an object, and at the end of each round participants must remember where they saw each object.

The Restoring Active Memory project run by the University of Pennsylvania has just released human intracranial brain recording and stimulation data for 102 new patients and a new spatial-navigation task developed by researchers at Columbia University. The total RAM dataset now includes such recordings from 251 patients and more than 1,100 experimental sessions, making it the largest publicly available dataset of its kind.

The ultimate aim of Restoring Active Memory, or RAM, is to develop a fully implantable device that can electrically stimulate the brain to improve memory function. The program’s immediate goal is to deliver new treatments for those who have experienced a traumatic brain injury, such as veterans returning from combat. In the long term, such therapies could also help patients with a range of ailments, from Alzheimer’s to dementia.

Along the way, the DARPA-funded project, led by Penn psychology professor Michael Kahana and researcher Daniel Rizzuto, director of cognitive neuromodulation, has made crucial strides in brain function and memory.

Earlier this year, the team published a paper in Current Biology showing for the first time that electrical stimulation delivered when memory was predicted to fail could improve memory function in the human brain. That same stimulation generally became disruptive when electrical pulses arrived during periods of effective memory function.

Michael Kahana and Daniel Rizzuto of the University of Pennsylvania lead the Restoring Active Memory project.

In addition to intracranial recordings from patients and hundreds of experimental sessions, the initial dataset from 2016 included neuro-anatomical information about electrode location, patient-behavior data and experimental-design documents. Cognitive tasks, chosen for their importance in carrying out activities of daily living, included two focused on free recall, one on paired-associate learning and one on spatial navigation.

For this data release, which offers that same information for the new patients, the researchers included results from a task called “Treasure Hunt.”

Joshua Jacobs, a professor at Columbia, likens it to a video game in which patients navigate around a desert island the size of a football field. When they reach a treasure chest, it opens to reveal an object, and at the end of each round participants must remember where they saw each object. It’s like a computerized version of the card game “Memory,” during which a player flips over a card and must find its match by remembering the partner card’s location.

“This task assesses their spatial memory,” Jacobs said. “We’ve been able to use intracranial recordings from this task to build models of human memory that can then be used to predict their performance. We can tell how well someone is doing on the task just by examining their brain recordings.”

All study participants recently underwent brain surgery at one of eight clinical centers across the country: Columbia, Emory and Thomas Jefferson universities; the Dartmouth School of Medicine; the Mayo Clinic; the University of Texas Southwestern Medical Center; the National Institute of Neurological Disorders and Stroke; and the Hospital of the University of Pennsylvania.

The RAM project, which began three years ago, aims to epitomize open science, a notion common in physics but less widespread in neuroscience. Offering the raw dataset and updates is a step in that direction.

“We’ve used these recordings to identify the neural biomarkers of human memory and to understand how stimulation influences brain physiology and behavior,” Kahana said. “Releasing these data publicly will allow other researchers to replicate our results and to discover new findings that will move the field forward.”

It’s proof-of-concept that it’s possible to successfully develop new brain-stimulation therapies for patients with memory disorders, Rizzuto added.

“This could lead, eventually, to great improvements for people with traumatic brain injury and Alzheimer’s disease,” he said.

To receive the raw dataset, interested researchers may request access through the RAM website. Funding for the Restoring Active Memory project comes from the Department of Defense’s Defense Advanced Research Projects Agency.