A method of ECG template extraction for biometrics applications

Area of Science:

• Biometrics
• Healthcare Technology
• Signal Processing

Background:

• Electrocardiogram (ECG) signals are crucial non-invasive physiological data in healthcare biometrics.
• ECG biometrics offer ease-of-monitoring, unique individual identification, and significant clinical value.

Purpose of the Study:

• To propose a novel dynamic threshold setting method for extracting stable ECG waveforms.
• To enhance the accuracy and efficiency of ECG identification processes.

Main Methods:

• Implemented a dynamic threshold setting technique to identify stable ECG waveform templates.
• Utilized dynamic time warping for difference measures in ECG biometrics.
• Evaluated the method on a self-built electrocardiogram database.

Main Results:

• Significantly improved the accuracy of ECG biometrics.
• Reduced the half total error rate from 3.35% to 1.45%.
• Achieved an average running time of approximately 0.06 seconds on Android mobile terminals, demonstrating real-time capability.

Conclusions:

• The proposed dynamic threshold method effectively enhances ECG biometric system accuracy.
• The method provides acceptable real-time performance, suitable for mobile health applications.
• This advancement contributes to more reliable and efficient ECG-based identification systems.