An open-access database of lead-I ECG signals with practical variability scenarios

The article addresses the growing need for secure and efficient biometric systems under scenarios with variability of ECG signal morphology in practical environments and evaluation of biometric systems with more than 500 individuals. Therefore, an open-access database that contains lead-I ECG signals is consolidated from an own database and mainly open-access public ECG databases to cover a wide diversity of intra-user variability scenarios such as physical activities, mental stress, physiological changes over time, acquisition postures, drug effects, and cardiac medical history. This consolidated database consists of 5620 lead-I ECG recordings for 562 individuals, where each individual has 10 variability observations of 10 seconds duration each. Furthermore, the database provides detailed information on gender, age, health conditions, acquisition posture, and life habits. In conclusion, this database represents a significant landmark in the field of biometrics and ambulatory cardiac diagnostics, opening new opportunities for research and development of more accurate, safe, and adaptive ECG signal processing systems.