| |
Abstract:
Numerous studies have shown a face-selective ERP (N170) and
MEG (M170) response occurring 170 msec after stimulus onset over
occipitotemoral sensors. Here we asked what aspects of a face
stimulus are critical for this M170 response. The magnitude and
latency of the M170 at pre-defined face-selective sensors of
interest (SOI) was measured for stimuli in which we orthogonally
varied whether the face contained I) Real face parts (eyes, nose,
mouth) versus solid black ovals in their corresponding locations;
II) Veridical face configurations versus rearranged nonface
configurations; and III) Intact external face features versus none.
Subjects passively viewed the stimuli in random order while
fixating. Our data indicate that any two components together can
elicit an M170 as large as that for normal faces; any component
alone, however, is not sufficient for a full-amplitude response. In
addition, external face features interact with other stimulus
information, potentiating a maximal response even to a nonface
configuration of ovals despite the fact that neither the ovals
alone nor the external face features alone produce a response
stronger than the control condition (houses). The overall MEG
response profile is very different from that seen in a parallel
fMRI study of the FFA using the same stimuli, suggesting that the
two measures tap into either different neural populations or
different phases of the response in the same neural
population.
|
