Localization of visually evoked cortical activity in humans

Abstract

The locations of cortical activity evoked by visual stimuli presented at different positions in the visual field are deduced from the scalp topography of visually evoked potentials in humans. To accomplish this, the Laplacian evoked potential is measured using a multi-electrode array. It is shown that the Laplacian response has the following useful attributes for this purpose. It is reference-free. Its spatial resolution is approximately 2 cm referred to the surface of the cortex. Its spatial sensitivity characteristic is that of a spatial band-pass filter. It is relatively insensitive to source--sink configurations that are oriented tangentially to the surface of the scalp. Only modest assumptions about the source--sink configuration are required to obtain a unique inversion of the scalp topography. Stimuli consisting of checkerboard-filled octant or annular octant segments are presented as appearance-disappearance pulses at sixteen different positions in the visual field in randomized order. The locations of evoked cortical activity in the occipital, parietal and temporal lobes are represented on a Mercator projection map for each octant or octant segment stimulated. Lower hemifield stimuli activate cortex which lies mainly on the convexity of the occipital lobe contralateral to the side of stimulus presentation in the visual field. The more peripheral the stimulus is in the visual field, the more rostral is the location of the active cortex. The rostral-to-caudal location of the evoked activity varies from subject to subject by as much as 3 cm on the surface of the occipital cortex. Furthermore, in any single subject there is a substantial amount of hemispheric asymmetry. Upper hemifield stimuli activate cortex that lies on the extreme caudal pole of the occipital lobe. This activity is relatively weak, and in some subjects it is almost unmeasurable. It is suggested that the representation of the upper hemifield in the cortex lies mostly on the inferior and mesial walls of the occipital lobe and possibly within the calcarine fissures. Those locations are inaccessible to the Laplacian analysis because the current generators therein may be oriented tangentially to the surface of the overlying scalp. Posterior parietal lobe activity and/or inferior temporal lobe activity is frequently evoked. Different subjects have different patterns of evoked activity. Unilateral or bilateral posterior parietal lobe activity is the most common pattern. Unilateral inferior temporal lobe activity is a less common pattern.(ABSTRACT TRUNCATED AT 400 WORDS)