Effect of Signal Conditioning and Evoked-Potential Based Representation on Stability and Distinctiveness of EEG Brain Signatures

Biometric patterns have been used for authentication purposes for more than decades. Although traditional biometric measurements including fingerprints, facial recognition, and sound recognition are often used for authentication purposes, there is ongoing concern about the risks associated with their repeatability and susceptibility to theft. The utilization of electroencephalography (EEG) has promising potential thanks to the distinctive patterns extracted from brain signals. The main goal of this study is to explore the effectiveness of signal conditioning and evoked potentials on uniqueness and permanence of individuals' EEG signals for authentication purpose. We utilized the Laplacian of Gaussian as a signal conditioning operator. In addition, three signifying functions were explored in obtaining evoked potential time-domain representation from an individual EEG signal. Different stimuli protocols were employed for comparison. A time-series and a statistical technique were utilized for uniqueness and permanence analysis.