Summary: Researchers have made a major breakthrough in neural prosthetics: They have demonstrated the ability to recall specific memories using a new memory decoding model (MDM). The breakthrough builds on previous research to decode and stimulate neural activity using surgically implanted electrodes in the hippocampus, enabling targeted memory retrieval.
A study of 14 adult epilepsy patients showed that MDM stimulation significantly improved memory performance, especially in people with memory impairment, offering hope for the treatment of diseases such as Alzheimer’s disease, stroke and head injury. The research represents an important step towards the development of interventions to restore lost memory functions and could contribute to more independent living.
Important facts:
- The neural prosthetic device uses a memory decoding model to enhance the retrieval of specific memories and has demonstrated its success in human participants.
- Through targeted electrical stimulation in the hippocampus, significant improvements in memory performance were observed, especially in participants with memory impairment.
- DARPA-funded research opens new avenues for treating memory loss associated with Alzheimer’s disease, stroke, and traumatic brain injuries, and aims for interventions that promote independent living.
Source: Wake Forest University
A team of scientists from Wake Forest University School of Medicine and the University of Southern California (USC) has demonstrated the first successful use of neural prostheses to retrieve specific memories.
The results are published online in Frontiers in Computational Neuroscience.
This important research follows a 2018 study by a Wake Forest and USC team led by Robert Hampson, PhD, professor of regenerative medicine, translational neuroscience, and neurology at Wake Forest University School of Medicine, which demonstrated the successful implementation of an artificial system that uses a person’s brain to recreate memories.
Earlier studies introduced an advanced electronic artificial system designed around a nonlinear multi-input multi-output (MIMO) framework. This mathematical model allowed researchers to simulate complex neural interactions with high precision, paving the way for more targeted interventions in brain function.
Using this system, scientists successfully modulated the firing activity of neurons within the hippocampus—an area of the brain crucial for memory formation. Their approach demonstrated the potential to influence cognitive processes by directly interacting with neural circuits responsible for encoding new experiences.
In this study, researchers developed a new model of the process that helps the hippocampus remember certain information. When the brain tries to store or retrieve information such as “I put out the fire” or “Where did I put my car keys?”, groups of cells within neuronal ensembles work together to store or retrieve that information.
Using recordings of the activity of these brain cells, the researchers created a memory decoding model (MDM) that allowed them to understand what neural activity is used to store different pieces of specific information.
The neural activity decoded by MDM was then used to create a pattern or code that was used to neurostimulate the hippocampus while the brain attempted to store this information.
Brent Roeder, PhD, a leading researcher in Translational Neuroscience at Wake Forest University School of Medicine, unveiled a groundbreaking neurostimulation method aimed at enhancing memory function. The technique represents a major leap forward in cognitive science, offering new possibilities for personalized brain interventions.
What sets this approach apart is its precision. Rather than applying a broad, one-size-fits-all strategy, the stimulation can be tailored to reinforce specific memories that hold personal significance to the individual. This finding opens the door to more targeted therapies that align with each person’s unique cognitive needs.
The team recruited 14 adults with epilepsy who participated in a brain mapping diagnostic procedure in which electrodes were surgically implanted in different parts of the brain to identify the origin of their seizures.
Participants underwent all surgical procedures, postoperative monitoring, and neurological testing at one of three sites included in the study, including Atrium Health Wake Forest Baptist Medical Center, Keck Hospital of USC in Los Angeles, and Rancho Los Amigo National Rehabilitation Center in Downey, California.
The team administered electrical stimulation to the MDM during visual recognition tasks to determine whether it could help participants recall images better. They found that this electrical stimulation led to significant changes in the ability to remember information. Significant differences in performance were seen in about 22% of cases.
When they looked specifically at participants with memory impairment who had both hemispheres of their brain stimulated, about 40 percent of them showed significant changes in their memory performance.
“Our goal is to develop an intervention that can restore memory loss after Alzheimer’s disease, stroke, or head trauma,” said Roeder. “We found that the most dramatic changes occurred in people with memory impairment.”
Ruder said he hopes the technology can be further developed to help people live independently by helping them remember important information, such as whether they have taken their medication or whether the door is locked.
“Although more research is needed, we know that MDM-based stimulation has the potential to significantly alter memory,” said Ruder.
The research builds on more than 20 years of preliminary research on memory codes by Sam Dedewyler, PhD, professor emeritus of physiology and pharmacology at Wake Forest University School of Medicine, Hampson (now a member of the Wake Forest Institute for Regenerative Medicine) and a team of USC, Baird Medicine, PhD, Dong Song, PhD.
In preclinical work, the same type of stimulation was used to restore and promote memory in animal models using a MIMO system developed at USC.
Financing: This research was funded by the US Defense Advanced Research Projects Agency (DARPA).
Abstract
Development of hippocampal neural prostheses to facilitate human memory encoding and retrieval of stimulus features and categories
Objective: This study presents a pioneering achievement in memory research: the successful application of fixed neural stimulation patterns tailored to distinct informational content. These patterns are generated from a computational model that translates the hippocampus’s spatiotemporal coding mechanisms into personalized memory cues, enabling targeted stimulation aligned with individual memory representations.
Perspective: Our research introduces a novel framework for understanding how the hippocampus encodes distinct memory elements. This model captures the dynamic spatiotemporal activation of neuronal ensembles, revealing how specific patterns of neural activity contribute to the successful encoding of targeted information into short-term memory.
To support this framework, we developed a memory decoding model (MDM) that simulates the activation profiles of neurons in the hippocampal CA3 and CA1 regions. This computational tool enables precise mapping of neural responses, offering deeper insight into the mechanisms underlying memory formation and retrieval.
It derives the activation pattern of CA1 and CA3 neurons that is applied during the encoding phase (sampling) of a human short-term memory task using a delayed match-to-sample (DMS) method.
Key findings: Electrical MDM stimulation of the CA1 and CA3 regions of the hippocampus during the rehearsal phase of DMS trials facilitated retention of DMS images during a delayed recognition (DR) task that also included non-DMS control images. Across subjects, the stimulation trials revealed significant changes in performance across the 22.4% of patient category combinations. These changes in performance were a combination of increases and decreases in memory performance, with increases in performance relative to decreases of approximately 2:1.
In a clinical study involving individuals with memory impairments, bilateral stimulation led to notable shifts in cognitive outcomes. Over one-third of the patient-category pairings—specifically more than 37.9%—exhibited measurable changes following the intervention, suggesting a meaningful impact on neurological function.
When analyzing the direction of these changes, the results leaned heavily toward improvement. For every instance of decline in memory performance, there were more than four cases showing enhancement. This strong positive ratio highlights the potential efficacy of bilateral stimulation as a therapeutic approach for memory-related conditions.
The change in memory performance depended on whether memory function was intact or impaired and whether the stimulation was bilateral or unilateral. Almost all of the performance improvements were observed in individuals with memory impairment who received bilateral stimulation.
Significance: These results demonstrate that memory encoding can be facilitated for specific memory content in patients with memory impairment, providing a promising approach for future implantable neural prostheses to enhance human memory.
