Are electroencephalogram (EEG) signals pseudo-random number generators?

Abstract

In this paper, we study if an electroencephalogram (EEG) signal can be treated as a pseudo-random number generator (PRNG). We approach this problem by calculating the frequency of both healthy and epileptic EEG signals, and we find that all EEG signals obey the Gaussian distribution with different standard deviations. However, under some transformation, we can treat EEG signals as PRNGs. Our transformation takes the sum of the least significant five bits of the EEG sample amplitude and then outputs the parity of this sum to a binary (0 or 1) sequence. Our binary sequences have passed nearly all NIST pseudo-random number tests except a few failures. This indicates that the EEG signals can indeed become a PRNG provided that it undergoes a transformation like ours.