Identification of the hemisphere activated by hemifield visual stimulation using a single equivalent dipole model.

Topographic amplitude distribution of the hemifield pattern visual evoked potential (PVEP) shows substantial intersubject variability. Many subjects have larger P100 amplitudes paradoxically over the hemisphere ipsilateral to the stimulated field, whereas others show larger responses over the stimulated hemisphere. The present study was designed to determine whether a single equivalent dipole model could correctly identify the field of stimulation, and therefore the hemisphere activated, under conditions in which the surface distribution is variable. Under conditions used in the present study, visual examination of the surface amplitude distribution of the P100 peak could not be reliably used to identify the hemifield that was stimulated. Of 28 hemifield PVEPs, obtained from 14 normal subjects, only 13 showed higher amplitude ipsilateral to the field of stimulation. Thus, neither examination of EP wave forms nor topographic maps provided an accurate means for determination of which hemifield was stimulated or which hemisphere was activated. The single equivalent dipole model correctly identified the stimulated hemisphere for 25/28 hemifield PVEPs. Orientation of the equivalent dipole accounted for much of the variability in surface amplitude distribution, with tangential orientations obtained in subjects with ipsilaterally predominant P100 surface topography. Although the dipole model improved identification of the stimulated hemisphere, dipoles were located anterior or inferior to the occipital lobe in some subjects. Results suggest that dipole modeling can provide useful information regarding the source of surface recorded potentials.