A deep learning-based authentication system utilizing ECG signals achieved 98.5%-99.5% accuracy and a 0.1% false acceptance rate, providing secure authentication for telehealth environments.
An innovative deep learning model using ECG signals for biometric authentication achieved up to 99.5% accuracy, offering a secure method for telehealth systems.
As telehealth services have become integral to healthcare applications; robust authentication mechanisms are critical for safeguarding sensitive patient data and services. Conventional authentication techniques including passwords and tokens are susceptible to theft and security breaches. This vulnerability highlights the need for alternative methods that offer improved security measures and ease of use. Biometric authentication, which leverages unique physical and behavioral traits, has emerged as a promising alternative. Among various biometric modalities, electrocardiogram (ECG) signals stand out because of their uniqueness, stability, and noninvasive nature. This study introduces an innovative deeplearning-based authentication system that utilizes ECG signals to enhance security in Internet of Things (IoT)-powered telehealth environments. The proposed model employs hybrid architecture, starting with a Siamese Neural Network (SNN) for dynamic verification, followed by a Convolutional Neural Network (CNN) for feature extraction, utilizing an optimized Sequential Beat Aggregation approach for robust ECG-based authentication. The system operates securely and adaptively, and performs real-time authentication without requiring human intervention. The research approach involved the acquisition and processing of electrocardiogram data from the ECG-ID dataset which encompassed 310 ECG individuals obtained from 90 individual subjects. This dataset provided a comprehensive set of samples for training and evaluation. The model achieved high authentication accuracy (98.5 %–99.5 %) and a false acceptance rate of 0.1 % with minimal computational overhead, validating its feasibility for real-time applications. This study integrates ECGbased authentication into telehealth systems, creating a secure foundation for safeguarding patient data. The innovative use of ECG signals advances secure and adaptable for a personalized remote health monitoring system development.
Mohamed Azab (Thu,) conducted a other in Telehealth authentication (n=90). Deep learning-based authentication system (SNN and CNN) was evaluated on Authentication accuracy and false acceptance rate. A deep learning-based authentication system utilizing ECG signals achieved 98.5%-99.5% accuracy and a 0.1% false acceptance rate, providing secure authentication for telehealth environments.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: