Instantaneous frequency maps, dipole models and potential distributions of pattern reversal-evoked potential fields for correct recognition of stimulated hemiretinae

Abstract

Lateral hemifield pattern-reversal visual evoked potential (PVEP) field data were evaluated using potential distributions, dipole modelling and distributions of Hilbert transformation-based instantaneous frequency in order to determine the stimulated hemisphere. Twenty channel records were collected from 35 normal volunteers in two laboratories using similar stimulus conditions (11–20.5° target, 60–75 min checks, 2/s reversal, 500 ms analysis epoch). P100 latency was determined in each average by the global field power maximum between 90 and 120 ms. Using the data from O1 and O2 at P100 latency, the stimulated hemisphere was identified by maximal potential or minimal instantaneous frequency on the stimulus-contralateral side, or, using the 20-electrodes data at P100 by the ipsilateral lateralization of the dipole model. Correct classification of the stimulated 70 hemiretinae was achieved by potential distribution in 44 cases, by dipole modelling in 54 cases and by instantaneous frequencies in 68 cases. Errors in the classification by potential distribution and dipole location were twice as frequent for decisions based on expected locations over the left than over the right hemisphere. This finding might be caused by the relatively larger size of the left occipital lobe. We conclude that a single value of instantaneous frequency which implies a massive data reduction can serve as a robust parameter for the characterization of the input conditions of hemifield PVEP (i.e. the stimulated hemiretina). It is more successful than potential distribution or dipole modelling, probably because instantaneous frequency incorporates considerably more information than the other two measures. It is suggested to explore instantaneous frequency as a parameter to recognize small retinal area stimuli in perimetry studies.