39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society
Abstract
To specify reliable features for advanced electrencephalography (EEG) based biometrics, we extracted the power and network parameters (Eigenvector centrality: EC) using two types of phase coupling (mean phase coherence (MPC) and phase lag index (PLI)) in each of 4 distinct frequency bands. A Euclidean distance-based authentication algorithm demonstrated that the MPC of high (>20 Hz) frequency rhythms yielded the best performance (accuracy of 95.4%) and the equal error rate was 8.1%, and the accuracy rate was 97.8% when fusing all the features.