EEG-based biometrics: phase-locking value from gamma band performs well across heterogeneous datasets

Abstract

The performance of functional connectivity metrics is investigated for electroencephalogram (EEG)-based biometrics using a support vector machine classifier. Experiments are conducted on a heterogeneous EEG dataset of 184 subjects formed by pooling three distinct datasets recorded with different systems and protocols. The identification accuracy is found to be higher for higher frequency EEG bands, indicating the enhanced uniqueness of the neural signatures in beta and gamma bands. Using all the 56 EEG channels common to the three databases, the best identification accuracy of 97.4% is obtained using phase locking value-based measures extracted from the gamma frequency band. When the number of channels is reduced to 21 from 56, there is a marginal reduction of 2.4% only in the identification accuracy. Additional experiments are conducted to study the effect of the cognitive state of the subject and mismatched train/test conditions on the system performance.