Enhanced Information Security with ECG Biometric Authentication

The emergence of electronic health technologies has revolutionized the healthcare industry, leading to significant improvements in medical services at reduced and affordable costs. However, management of health information suffers from a lot of security and privacy challenges, which include user authentication, data integrity, data confidentiality, and safeguarding patient privacy. Biometric technologies address these issues by providing security model to ensure that only authorized personnel have access to their respective health data. This paper proposes an authentication approach utilizing Electrocardiogram (ECG) signals to enhance privacy and information security through the application of AC/DCT for feature extraction and using k-NN and SVM models as the classifiers. The study has been conducted on the NITH Multimodal Biometric Database, which comprises recordings of 15 healthy individuals and is divided into two sessions of data. The best performance on the basis of ACC and EER recorded on the dataset are 99% and 0.3% for same session data, and an ACC and EER of 96.4% and 1.19% are achieved for across session data.