fMRI Scans Shed Light on Mechanisms of Memory

Why is it we can remember every word of an annoying commercial but not where we last saw the television remote? The answer may lie in a study using magnetic resonance imaging (MRI) to capture what happens in the brain the instant a memory is formed.

The study, carried out by Anthony Wagner, a post-doctoral fellow in psychology, suggests that whether we remember something or not is related to how deeply we consider its meaning during memory formation.

"Thinking about an event's meaning yields on average a better memory," Wagner says. "It may be the subtle differences--the extent to which we think upon its meaning, relate it to our past and so on-that determine an event's memorability."

The study, carried out by researchers at Harvard, Boston's Massachusetts General Hospital and St. Louis's Washington University, scanned test subjects under two different learning conditions.

In one case, volunteers were rapidly shown words and asked to determine if they were abstract or concrete. In the other situation, test participants were asked about the physical form of the word--whether they were upper or lower case letters. In both instances, no one was told to memorize anything. Twenty minutes later, the subjects were shown groups of words and asked if they had seen them during the test.

The results were less-than-surprising to the scientists. Those who concentrated on the meaning of the words had, on average, a greater memory of what they'd seen. Even more revealing: the MRI scans showed greater activity, as measured by blood oxygenation level, for single items that were remembered as opposed to those that were forgotten.

"We asked what the brain was doing differently during experiences which were remembered," Wagner says. "The scan showed greater activation in the left frontal and left temporal region when people paid attention to the semantic meaning of words."

The left frontal region is the area of the brain behind the temple. The left temporal lobe is the area behind that. Previous neuroimaging studies have identified the parahippocamal cortex, the inner wall of the left temporal region, as particularly important in memory formation. Due to advances in MRI techniques, the activated regions can now be localized with greater precision than previously possible.

Does this mean that the left frontal and left temporal regions of the brain are responsible for memory formation? A similar study done by James Brewer of Stanford University showed that different areas of the brain are responsible for remembering (and possibly retrieving, although that was not studied) different types of events. In Brewer's experiment, subjects were shown pictures, not words, and greater activity was concentrated in the right frontal and right temporal regions of the brain.

"The studies are quite complimentary," Wagner says. "What we've jointly demonstrated is that both the frontal and medial temporal regions determine memory. But the specific structures employed depend upon the content and the approach to learning."

The use of MRI scans to isolate particular regions of the brain responsible for memory encoding has important implications for medicine. An early symptom of Alzheimer's disease, for example, is the inability to learn and remember new experiences.

"One possibility is to use this to test older adults to see if small changes [in these regions] are predictive of memory loss," Wagner says. "Once we have that, early diagnosis and early treatment [of Alzheimer's] are possible."

In general, Wagner's technique poses new questions about the aging process in healthy, normal adults. Although older adults can obviously remember new experiences, there is some impairment of how much is remembered.

"There's some imaging data suggesting that during learning they tend not to recruit [the same areas as younger adults,]" Wagner says. "Is this a strategic shift? If not, what can we do to recruit those areas?"

The two studies lend a neurological basis to what cognitive science has known for the past 100 years-that it is more effective to master a new skill by relating it to your knowledge as a whole than by rote repetition.

Wagner uses the example of learning vocabulary words in a foreign language.

"Each subsequent recital yields less and less activation of these areas of the brain. The best way to learn it is to space out learning trials."

Speaking of his own future research possibilities, Wagner says, "We want to find out why that is."