ABSTRACT
Our previous research in patients with extensive surgical ablations of the prefrontal cortex contradict the hypothesis of some authors that the generators of several auditory event-related potentials (ERPs) (N100; P200; N200; P300; SW), recordable in humans with depth/scalp electrodes and MEG over the prefrontal dorsolateral cortical areas, are essentially located in medial prefrontal and anterior cingulate-limbic cortices. Using a standard CNV paradigm, 21 EEG electrodes and topographic mapping analysis, the post-warning (S1) auditory N100a b c, P200, P300 (binaural clicks) and CNV activity were recorded in three additional patients after extensive dorsolateral and/or medial prefrontal cortex ablations, verified through CT/MRI examinations. No true post-S1/CNV components were recordable over the ablated frontal areas, only sporadic volume-conducted ERPs probably generated in the temporo-parietal lobes or posterior cingulate gyrus. For one of these patients, after excision of a vast right frontal epileptogenic cortical region (including extensive dorsolateral areas, but sparing the fronto-medial cortex and anterior/middle cingulate gyrus), no post-S1/CNV components were recordable over the ablated regions. These latest observations again indicate that independent neuronal generators of several post-S1 auditory and CNV components are located in the dorsolateral supramodal premotor/prefrontal cortical areas which are directly, ipsilaterally connected to the uni/multimodal temporo-parieto-occipital sensory and associative regions through the long, two-way, fairly superficial, superior arcuate-longitudinal and deeper superior and inferior occipito-frontal bundles. Clear and almost constant differences in the latency of some post-S1 N100 subcomponents (especially the time-lapses between onset and the highest amplitude of the N100 a and c) over various posterior, central and anterior cortical areas sequentially involved, roughly measured in 10 normal subjects along the scalp and with MRI cerebral imaging, may probably be accounted for by the transcortical homohemispheric conduction time, which varies in our scalp recordings from 1 cm/0.74-1.28 ms, mean approximately 1 cm/1.02 ms ( approximately 9.8 ms).
