Memorable research

PU team devises way to track strolls down memory lane

By: David Campbell
   Neuroscience researchers at Princeton University have developed a new method for tracking people’s mental states as they remember past events.
   In the process, they may well have found scientific proof of Marcel Proust’s famous madeleine cookie — or Tony Soprano’s Proustian slice of cappicola, depending on how you take your take your pop wisdom.
   The memory research was the dissertation work of Sean Polyn, who earned his doctorate in psychology from Princeton in 2005 and is a now a postdoctoral researcher at the University of Pennsylvania.
   The project was conducted in the lab of Professor Kenneth Norman, an assistant professor of psychology at Princeton. Psychology Professor Jonathan Cohen, director of the university’s Center for the Study of Brain, Mind and Behavior, and Vaidehi Natu, a researcher in Professor Norman’s lab, also collaborated.
   The findings, detailed in the Dec. 23 issue of the journal Science, advance the understanding of the brain’s memory systems, and could lead to applications in the study and treatment of brain disorders like Alzheimer’s disease, the researchers said.
   The researchers showed nine participants a series of pictures, then asked them to recall what they had seen. Subjects’ brain activity was monitored the whole time using Princeton’s functional magnetic resonance imaging or fMRI scanner, located in the basement of the university’s Green Hall, and analyzed using a computerized pattern-recognition program.
   In the study, which was funded with grants from the National Institute of Mental Health, participants studied and then were asked to recall a total of 90 images in three categories — celebrity faces, famous locations and common objects. Among the images: Halle Berry, Jack Nicholson and Carrot Top; the Taj Mahal and the pyramids at Giza; Legos and tweezers.
   The researchers used the computer to distinguish among the patterns of brain activity associated with studying each of the three categories, then to track participants’ brain activity as they recalled the images to see how well it matched the patterns associated with the initial viewing of the images.
   They found that patterns of brain activity for specific categories, such as faces, started to emerge about five seconds before the subjects recalled specific items from that category, suggesting that participants were cueing specific memories by first bringing to mind the general properties of the images.
   "As people tried to think back, the brain would get gradually more and more like what it was when they were studying the categories," Professor Norman said.
   "People’s brains started to get ‘face-like’ on average around five seconds before they remembered a specific face," he said. "They would become more ‘location-like’ about five seconds before they remembered a specific location."
   The researchers showed that participants’ brain state gradually aligned with their brain state from when they first studied the pictures, supporting the theory that memory retrieval is a kind of "mental time travel," a term coined by Professor Endel Tulving of the University of Toronto.
   Also, by measuring second-by-second changes in the brain during the process of remembering, the researchers were able to predict what kind of item the subjects would recall next.
   Mr. Polyn said the study examined a memory paradigm called "free recall" — that is, participants were asked to remember whatever they wished and in no particular order.
   He said people searched for specific memories using a "bootstrapping" process, in which they used a piece of the desired memory to pull up the whole thing.
   But he said the study’s findings can also apply to cases of unintentional recall, such as in Proust’s "In Search of Lost Time," in which the taste of a cookie dipped in tea prompts unbidden memories of the narrator’s childhood.
   "If there’s one thing memory researchers have figured out, it’s that what you get out of memory depends on how you prompt memory," Professor Norman said. "The main determinant of how you get back to something is — do you have a good reminder."