Silent speech activates prefrontal cortical regions asymmetrically, as well as speech-related areas in the dominant hemisphere.

Regional cerebral blood flow (rCBF) was measured simultaneously over the right and left hemispheres by 2 x 32 detectors in 30 healthy volunteers with a two-dimensional iv xenon-133 technique, during (1) rest and during (2) audible and (3) silent counting (101, 102,...). Mean hemisphere CBF increased significantly in both hemispheres during the activations. Audible speech activated rolandic and temporoparietal regions mainly on the right side. This pattern covers auditory and para-auditory as well as motor (tongue/larynx) regions. Most likely those regions are involved in auditory feedback and voice control. Silent speech (internal speech) gave a clearly different activation pattern involving (1) left-sided regions related to speech perception and speech motor control (including the SMA) and (2) a right dorsolateral prefrontal area that may be related to attention mechanisms. The silent speech pattern appears to demonstrate aspects of internal (cognitive) feedback activity in which prefrontal cortical regions are activated significantly. Audible and silent counting may represent two principally different types of cerebral feedback systems, one for overt sensory-motor activity and one for a pure internal cognitive feed-back.