On the use of the electrocardiogram for biometrie authentication

This paper provides a comprehensive survey of the deployment of the electrocardiogram (ECG) as a biometrie authentication system. The paper begins with an overview of the ECG, some of the features that have been extracted for use in biométrie authentication, followed by a series of case studies. The conclusion reached in this survey indicates that the ECG is a very effective and efficient mode of user authentication. Classification accuracies of over 95% are routinely obtained, rivalling more traditional anthropométrie based approaches such as fingerprint and face recognition based approaches. Some caveats of the ECG based approach are also discussed — these include accounting for disease states, emotional states, and the general caveat that many of the published studies are performed on a small cohort.