Blockchain-Enabled ECG Biometrics for Secure Identity Management

Abstract

Fingerprints, facial features, and irises are all examples of classic biometric data, and it is feasible to either copy or intercept this type of data. We are introducing a decentralized identity technique that utilizes ECG patterns of the heart as a biometric that is more difficult to fake in order to address this issue. Smart contracts are employed to encrypt and preserve the templates that are produced on a private blockchain, whereas deep learning networks are responsible for managing feature extraction. This eliminates the single centralized point of vulnerability and improves the reliability of credentials. The trials that were conducted with PhysioNet's MIT-BIH Arrhythmia and PTB-XL datasets resulted in error rates (EERs) of 0.08% and 0.11%, respectively. These results are an indication that the performance of the authentication process was strong.