ECG biometric recognition: Permanence analysis of QRS signals for 24 hours continuous authentication

Recent studies regard the use of ECG signals for biometric recognition exploiting the possibility of these signals to be frequently recorded for long time periods without any explicit actions performed by the users during the acquisitions. This aspect makes ECG signals particularly suitable for continuous authentication applications. In this context, researches have proved that the QRS complex is the most stable component of the ECG signal. In this paper, we perform a preliminary study on the persistency of QRS signals for continuous authentication systems. A recognition method based on multiple leads is proposed, and used to evaluate the persistency of the QRS complex in 24 hours Holter signals. This time interval can be considered as adequate for many possible applications in continuous authentication scenarios. The analysis is performed on a significantly large public Holter dataset and aims to search accurate matching and enrollment strategies for continuous authentication systems. At the best our knowledge, the results presented in this paper are based on the biggest set of ECG signals used to design continuous authentication applications in the literature. Results suggest that the QRS complex is stable only for a relatively small time period, and the performance of the proposed recognition method starts decreasing after two hours.