| |
Abstract:
Functional neuroimaging studies have associated old/new
recognition memory with neural activity in frontal and medial
temporal regions. Are these activations specific to old items or do
they also occur for new items? Earlier functional neuroimaging
studies investigated this problem by presenting old and new items
in different test blocks, at the risk of altering the retrieval
strategies for the two types of items. Recently, event-related fMRI
paradigms allow for the presentation of old and new items in a
mixed test sequence, as in standard cognitive experiments. We used
event-related fMRI to investigate the neural correlates of
veridical and illusory recognition. Before each fMRI run,
participants watched a videotape in which a male and a female
speaker alternatively read six lists of words. The words in the
lists were either converging associates to a nonpresented critical
lure or instances of a natural category. During fMRI scanning,
subjects were presented with (1) words presented in the lists (TRUE
items), (2) new words closely related to studied words (FALSE
items), and (3) new unrelated words (NEW items). Preliminary
analyses on the results of nine subjects indicated that, although
new items elicited frontal activation relative to baseline,
posterior frontal and medial temporal regions were more activated
for TRUE than for NEW items. Interestingly, a medial temporal
region was also identified in the FALSE-NEW contrast. These results
suggest that activation in some frontal and medial temporal regions
is greater when retrieval is successful. At the same time, medial
temporal activity may also underlie false recognition.
|
