| |
Abstract:
Abstract: We used event-related fMRI to measure brain
activity while subjects performed an item-recognition
working-memory task in which the memory-set size varied between 1
and 8 letters. Each trial was composed of (1) a 4-second encoding
period in which subjects viewed random letter strings, (2) a
12-second retention period and (3) a 2-second retrieval period in
which subjects decided whether a single probe letter was or was not
part of the memory set. In the aggregate data, monotonic effects of
memory-set size on fMRI signal intensity were observed in the
retention period and retrieval periods in dorsolateral prefrontal
cortex. The direction of these effects varied with subjects'
overall task performance. Good performers showed negative monotonic
effects at encoding and positive monotonic effects during retention
and retrieval periods. Poor performers showed positive monotonic
effects at encoding and retention, but no monotonic effects during
retrieval. Regression analyses of individual subjects' reaction
time and cortical activity indicated that speed and accuracy of
performance accounted for considerable variance in fMRI signal
intensity during encoding and retrieval in dorsal and ventral
prefrontal cortex. These results are consistent with a model of
cognitive slowing in which processing rate is related to neural
efficiency. The present results suggest that relationships between
neural activity and performance vary with increasing mnemonic
demand.
|
