Non-linear Analysis of Psychophysiological Effects of Auditory Stimuli using Fractal Mining

While spectral analysis (e.g., Fast Fourier Transformation) of electroencephalogram (EEG) has been one of the most well established parameters in psychophysiology, physiological implication of fractal analysis has not been established. Further, systematic examination of the association between the waveforms of auditory stimuli and EEG is an untouched area of research. In the present study, we used fractal analysis and data mining techniques, and created a method for finding the association between fractal dimensions of auditory stimuli and fractal dimensions of EEG. Applying our method, we found strong associations between signal complexity in auditory input and the resulting EEG data, with confidence values exceeding 90% in several of the associations. Our success in this initial application could potentially be generalized to further brain activity analysis.