| |
Abstract:
Abstract: The perceptual awareness of a given event is
vulnerable to many types of task manipulations. Backward visual
masking is one such phenomenon, in which the presentation of a
nonprimary image (the mask) can reduce or eliminate a subject's
ability to detect a primary visual stimulus presented earlier.
Under appropriate masking conditions, identical stimulus
presentations will result in trial-by-trial differences in the
subject's awareness of the stimulus. Event-related fMRI was used to
determine the pattern of neural activation correlated with this
awareness. Subjects were shown brief presentations of INTACT,
SCRAMBLED, or MASKED faces (an intact face masked by a subsequent
scrambled face). Activation of the fusiform face area (FFA) was
then compared across trial types. Peak activation was significantly
greater for INTACT trials than for SCRAMBLED trials, with the
activation for MASKED trials falling between these extremes. When
the MASKED trials were segregated post hoc according to the
subject's perception, a significantly greater activation was found
for trials in which a face was detected. A further segregation of
the MASKED trials yielded a greater activation for those trials in
which the face was incorrectly identified. These findings indicate
that the activation of inferotemporal cortex, and the FFA in
particular, is correlated with the subject's perception of the
visual image rather than the contents of the image per se. (NSF
BCS-9996264)
|
