Both Hermite polynomial and CWT-based deep learning approaches provide highly accurate ECG-based biometric identification, with a trade-off between computational efficiency and robustness.
ECG biometrics merit security exploration; leaves open prospective validation in diverse populations.
This study presents a novel framework for biometric identification using electrocardiogram (ECG) signals, using Hermite polynomial coefficients and Continuous Wavelet Transform (CWT) with convolutional neural networks (CNNs). The first approach models ECG beats via Hermite polynomials, achieving 98.33% accuracy on the MIT-BIH6 database, while the second leverages CWT-derived scalograms (complex Morlet wavelet) to reach 98.50% accuracy. A systematic comparison reveals key trade-offs: Hermite-based 1D CNNs excel in computational efficiency (59% faster training), whereas 2D CNNs with CWT scalograms offer superior robustness. These results demonstrate the viability of ECG as a biometric modality and provide a benchmark for future research.
No takes yet. Share an insight, caveat, or question.
Amira (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: