| |
Abstract:
We obtained fMRI scans while healthy volunteers performed
either a direct or indirect memory task on famous and nonfamous
faces, each of which was presented twice in a randomly intermixed
design. In the direct episodic task, volunteers indicated with a
keypress whether it was the first or second time they saw each
face, regardless of whether or not it was famous. In the indirect
semantic task, volunteers indicated whether each face was famous,
regardless of whether or not it was seen before. Reaction times
were faster to second than first presentations in the indirect task
(a priming effect), but not in the direct task. BOLD event-related
response decreases associated with repetition of faces were
observed in bilateral fusiform regions in the indirect but not
direct task (mirroring the reaction times). Repetition-related
response increases were observed in bilateral anterior and dorsal
prefrontal regions, and medial and bilateral parietal regions
(regions often associated with successful episodic retrieval) in
both tasks. Counter to expectations, the repetition-related
increases in most of these regions (except anterior prefrontal
cortex) were greater in the indirect than direct task. These
results suggest that 1) the BOLD repetition suppression in
extrastriate regions previously attributed to perceptual priming
may not be automatic (ie, task independent), 2) the
prefrontal/parietal increases previously attributed to episodic
retrieval success can be greater in indirect than direct memory
tests.
|
