Cohen's work opens up the world of memory

By looking at what goes wrong in the memory systems of amnesic patients, Neal Cohen's Amnesia Research Laboratory at the Beckman Institute has provided original insights into how memory functions in a healthy brain.

Neal Cohen's research into how memory is organized in the brain has relied heavily on those for whom remembering is a struggle. By looking at what goes wrong in the memory systems of amnesic patients, Cohen's Amnesia Research Laboratory at the Beckman Institute has provided original insights into how memory functions in a healthy brain.

For 16 years, the lab has produced seminal research showing that multiple, separate memory systems in the brain - particularly one system that critically includes the hippocampus - are responsible for much of everyday memory. How we remember, Cohen says, occurs through a process that "binds together the constituent elements of day-to-day events into enduring representations of experience."

During those 16 years the contributions of the amnesic patients, who have suffered brain damage through injury or disease, have been an invaluable part of the research. And recently it was discovered that his research could, in the future, aid victims of amnesia.

"The patients who have been kind enough to work with us have done so despite it being largely a basic research operation because they feel something good should come out of their misfortune," Cohen said. "But more recently we've been trying to use some of these methods on a case-by-case basis to develop some skills and to enhance the interaction of a patient with their caregiver."

As part of the lab's work, researchers found that patients and their caregivers can develop what is called common ground, or a shared knowledge, that enables more efficient communication. With common ground speakers don't have to explain the meaning of every word or phrase. During their studies Cohen and his fellow researchers learned that amnesic patients and their caregivers could, through repeated interaction, come up collaboratively with more concise, efficient ways to describe the objects with which they were interacting.

"What we found was that it's very powerful," Cohen said. "They did extraordinarily well at developing this common ground and we thought, ah, this is something we can use in the clinical applications to help people."

Research that may benefit amnesic patients is in its earliest stages, and it is not the focus of the lab's efforts. But it does offer the potential of adding another component to the work - one that could help many people in the future.

"The clinical research implications are that if you know what the strengths and weaknesses are, you can emphasize the strengths and then lead them to those areas of memory that are preserved so that they may make use of that in their day-to-day life," Cohen said.

"The research implications of that have to do with the explanation of the phenomena."

Explaining the phenomena is at the heart of Cohen's research. His Ph.D. focus at the University of California, San Diego, was on amnesia and multiple memory systems. After stops at MIT and Johns Hopkins, Cohen joined the Psychology Department at the University of Illinois at Urbana-Champaign. He came to the Beckman Institute in 1990.

Cohen's research lines have broadened and deepened since 1990, but his work is still centered on understanding memory by learning about the effects of amnesia. He points out that real amnesia is more complex and more interesting than typically portrayed in Hollywood movies, and he teaches a course at the university (Memory and Amnesia) on this topic.

"Everything we do springs from one central feature of our work, which is that despite profound memory impairment in amnesia there are aspects or domains that are fully intact," Cohen said. "So no matter how profoundly these patients are impaired following brain damage, there are preserved memory abilities. That has clinical implications, basic research implications, and applied implications."

The basic research implications are that those fully intact domains in the brains of amnesia patients help to show the existence of multiple memory systems.

"How is it that, following brain damage that prevents you from remembering your day-to-day experiences, it is possible for you to nonetheless be able to acquire new skills?" Cohen said. "It's an amazing thing; these people acquire skills as well as you and I do, but they don't remember the events and experiences during which those skills emerged. How do they do that?

"Our proposed solution to that is that there are multiple memory systems in the brain. That idea is now a commonly accepted explanation of the organization of memory in the brain. There are different systems that are functionally and anatomically separate, and you can have damage to one or the other and see the operation of the ones that are left intact."

Cohen said the concept of multiple systems is now the dominant view of memory organization in the field of neuroscience. His thesis work showed that preserved learning extends beyond the motor domain, and it differentiated between what is now known as declarative memory and procedural memory. The lab's work over the years has bolstered those views.

"Our particular take on what the hippocampus does in the brain is that the hippocampus, the system damaged in amnesia, is for declarative memory - memory for the relations among things - whereas non-hippocampal [systems are] more about procedural memory, the ability to support skills."

The lab's ability to characterize those memory systems in detail and demonstrate how the declarative system is about the binding of the relations among elements make it one of the leading facilities of its kind in the country.

Cohen, a member of the Cognitive Neuroscience group at Beckman, has co-authored with Howard Eichenbaum two important books in the field, Memory, Amnesia, and the Hippocampal System (MIT Press) and Memory Systems of the Brain (Oxford University Press).

In addition to the basic research, an important part of the lab's efforts are the clinical and applied aspects of the work, including such things as studying and developing rehabilitation strategies and refining the methods used for testing. The use of eye-tracking technology is a key feature of the lab's testing.

"It turns out that the way in which you move your eyes over a visual display or over a visual world is shaped by experience," Cohen said. "Then we can use the pattern of eye movements to tell us something about what you've experienced before, what things you've seen and not seen before, what do you know about this scene or not know about this scene."

Studying rehabilitation strategies for amnesic patients has implications for both applied and clinical research, Cohen said.

"There's two pieces to the research," he said. "One is on the applied end. We can use eye movements as a very powerful and robust measure of memory independent and separate of a verbal report. One way to say it is the eyes know even when we don't.

"On the more clinical end, the fundamental contribution of our lab year in and year out is the identification of aspects of preserved learning and memory in these patients, and our ability to use it both toward rehabilitation and toward an understanding of how memory is organized in the normal brain."

One possible application of the research could include helping in eyewitness identification cases by using eye movements as the measure of memory rather than using verbal reports. Applying the findings toward rehabilitation strategies for amnesic patients is potentially useful. But, Cohen said, there is not a proven rehabilitation strategy for amnesia and doing memory exercises hasn't proven very helpful either, although the use of mechanical devices such as PDAs can be beneficial.

"So it seems to us there's only two strategies: learn how to use external memory aids and/or learn how to take advantage of those aspects of memory that are intact and apply them, make them more useful, broaden the scope of the things that those memory abilities can do for you," Cohen said.

But he is hopeful the strategies that come out of this work will prove useful.

"We think it's potentially very valuable, and a recent graduate of our lab, Melissa Duff (now at the University of Iowa College of Medicine), is pushing this very hard," Cohen said. "We're very optimistic."

While the rehabilitation line of research has its merits, the lab's focus will still be on basic research.

"It's also incredibly gratifying that we can use these studies to learn about the normal condition," Cohen said. "There's no doubt that we've learned an immense amount about how memory is organized in the normal brain by seeing these cases, by inferring organization from how memory breaks down following brain injury."

Cohen has several ongoing collaborations with different faculty members, including a recent productive one with Cognitive Science group members Kathryn Bock and Gary Dell involving language production. The results showing that amnesic patients exhibited a speech characteristic called structural priming was a finding that reinforced both Cohen's theory of multiple memory systems and Bock's theory of implicit learning.

Cohen said that their collaborative work was predicated upon and provided a test of both their theories. "Which is lovely. That's what these collaborations are about."

Cohen said that an increasingly diverse number of collaborations and the growing use of advanced technologies such as eye-tracking and magnetic resonance imaging are currently producing an explosion of innovative research in the field of neuroscience, and in his own research area.

"It's a propitious moment right now as other work converges with our own to show that our characterization of what this system does for a living seems to be correct," he said. "So we're pushing out in a number of directions. One of the exciting things is we believe that this kind of memory works through the interaction of hippocampus and other systems."

Cohen said memory is stored in various areas of the cortex, not in the hippocampus, and the hippocampus serves to bind together in memory each of the different features each in their own processing area. So spatial information, face information, and language information each are stored in the areas that do the processing for that particular function.

"The hippocampus helps to bind it together and establish those connections in the cortical sites," Cohen said. "Only now, and I mean 2006, do we have the ability to look at the interactions among systems, both with MRI and optical imaging (near-infrared imaging). Both provide us with a means to look at those interactions here and elsewhere. That's an incredibly exciting thing."