Optimizing the steady-state multifocal visual evoked potential

Aims: The multi focal Visual Evoked Potential (mfVEP) is used in clinical diagnosis and basic research. Employing steady-state stimuli rather than the usual m-sequence stimuli may simplify statistical evaluation and reduce examination durations. This study aims to optimise performance (SNR) of the steady-state multi focal visual evoked potential (ssmfVEP) by examining stimulus characteristics such as size and the use of sparse stimuli.Methods: Normal adult human subjects (aged 18-60 years, n=41) were examined. SsmfVEPs using horizontal oriented sinusoidal luminance profile gratings were compared with conventional m-sequence mfVEP. In experiment one, stimuli at 48% contrast were modulated simultaneously in either 9 or 17 visual field regions comprising three rings presented at viewing distances of 16, 32 and 128cm. In experiment two, stimuli at 97% contrast were in two concentric regions, subdivided into quadrants and their SNR assessed for sparse and full field presentations. Three evoked potential channels were recorded, and three derived Lp-norm channels were calculated from them and assessed for efficient pooling.Results: In experiment one, SNR was significantly larger for m-sequence mfVEPs than for ssmNEP in all six channels. There was no effect of viewing distance for both ssmfVEPs and m-sequence mfVEPs. The larger regions (9 regions) gave larger responses than the small ones (17 regions). In experiment two, increasing both the number of concurrently stimulated regions and their proximity was associated with increased SNR. In both experiments, the norm channels gave the largest responses, indicating pooling was achieved.Conclusions: In normals, SNR of the ssmfVEP was unaffected by viewing distance and not improved by sparse stimulation. SNR was improved by increasing stimuli size. Future studies in glaucoma patients would be useful. The major findings provided evidence spatial summation occurred for all stimulus regions as the response increased as a function of increasing the number of concurrently stimulated regions, and that the degree of brain folding was similar for the three viewing distances. Conventional mfVEPs gave higher absolute SNR but lower SNR per unit time than ssmfVEPs in this study. The norm channels were more efficient than the individual channels