| |
Abstract:
Synthesized vowels, (/a/ and /u/) with male and female
pitches, were presented while neuromagnetic fields were recorded.
Participants were asked either to listen passively, to count
occurrences of a pitch (male), or to count the occurrences of a
vowel (/a/). The magnetoencephalographic (MEG) fields to binaurally
presented vowels were compared to responses to 1 kHz tones. Brain
sources of the magnetic fields were modeled as single equivalent
current dipoles (ECDs), and the magnetic field sources for both
vowels and tones localized to the supratemporal auditory cortex.
Tones showed no significant asymmetry in the number of ECDs
modeled. Vowels showed more than twice as many ECDs in the left
hemisphere compared to the right (p < 0.001). The anatomical
distribution of ECDs to vowels in the left was more restricted in
XYZ space than in the right hemisphere where ECDs were more widely
dispersed. A power analysis of MEG RMS did not reflect the dipolar
asymmetry. The task did not elicit strong attention effects. The
MEG data suggest an early, lateralized cortical mechanism for vowel
processing that is dipolar and independent of power. Further the
data suggest the number of MEG ECDs might be used as noninvasive
method for identifying hemispheric language dominance.
|
