| |
Abstract:
Abstract: Prefrontal cortex has been implicated in a variety
of higher cognitive functions from working memory to reasoning, but
how components of these functions are mapped to prefrontal cortex
remains unresolved. One possibility is that some regions serve to
maintain simple information, others are recruited as a
representation taxes capacity, and still others are specialized for
meaningful structure in a representation. To test these ideas, we
presented subjects with a delayed match-to-sample task. The sample
consisted of shapes distributed among six boxes. After a delay, a
probe screen containing seven shapes was presented. Subjects
responded "yes" only when the probe contained all the shapes in one
or more boxes. Trials with one shape per box or shapes in only one
box comprised, respectively, simple disjunctive and conjunctive
memory load tests. When multiple boxes contained multiple shapes, a
logical structure of combined conjunction and disjunction relations
had to be represented. By using two, four, or six shapes in the
sample we varied memory load and, in the structural trials,
complexity. During 98 trials 20 axial slices were acquired
(TR=2000, TE=35, flip=65, thickness=3.8mm) with a 1.5 Tesla MR
scanner in 13 subjects. An items X complexity X trial epoch group
ANOVA revealed that, while some prefrontal and parietal regions
responded linearly to load, other areas in frontal cortex responded
differentially when structure increased or when load approached
capacity.
|
