Dimensional Complexity Distribution on Scalp of Visual Evoked Potentials

Nonlinear dynamic properties in evoked potential (EP) with correlation dimension (CD) and the distribution of CDs of EPs on scalp are studied. EPs are recorded by averaging method. The CDs of EEG before and after stimulations (averaged) are calculated by Grass Berger and Procaccia method. The CDs of EP in multi-channel and their distribution are calculated. Results from this study show that the CD of evoked potentials in each channel is lower than spontaneous EEG. EP signals have less complicated properties than spontaneous EEG. CDs of EP in central back Oz positions are bigger than that in the central front and in side positions, it seems EPs around the source in occipital have more complexity. But by comparing the differences between CDs of EP and spontaneous EEG, EPs reduce the dimension more in source area than in other areas.