Key result
Transformer-based sequence pair extractor achieves ~93% peak ECG identification accuracy over long time separations.
Why the study?
Existing ECG biometrics perform poorly when ECG changes after enrollment because feature extraction cannot relate ECGs collected during enrollment to those collected during classification.
A transformer-based sequence pair feature extractor improves ECG biometric identification accuracy, particularly over long time separations between enrollment and classification.
Answers are grounded in this paper and cite their sources.
May advance ECG biometrics for identification; hypothesis-generating and requires prospective validation before clinical use.
The existing electrocardiogram (ECG) biometrics do not perform well when ECG changes after the enrollment phase because the feature extraction is not able to relate ECG collected during enrollment and ECG collected during classification. In this research, we propose the sequence pair feature extractor, inspired by Bidirectional Encoder Representations from Transformers (BERT)'s sentence pair task, to obtain a dynamic representation of a pair of ECGs. We also propose using the self-attention mechanism of the transformer to draw an inter-identity relationship when performing ECG identification tasks. The model was trained once with datasets built from 10 ECG databases, and then, it was applied to six other ECG databases without retraining. We emphasize the significance of the time separation between enrollment and classification when presenting the results. The model scored 96.20%, 100.0%, 99.91%, 96.09%, 96.35%, and 98.10% identification accuracy on MIT-BIH Atrial Fibrillation Database (AFDB), Combined measurement of ECG, Breathing and Seismocardiograms (CEBSDB), MIT-BIH Normal Sinus Rhythm Database (NSRDB), MIT-BIH ST Change Database (STDB), ECG-ID Database (ECGIDDB), and PTB Diagnostic ECG Database (PTBDB), respectively, over a short time separation. The model scored 92.70% and 64.16% identification accuracy on ECGIDDB and PTBDB, respectively, over a long time separation, which is a significant improvement compared to state-of-the-art methods.
No takes yet. Share an insight, caveat, or question.
Chee et al. (2022) studied ECG biometrics. Sequence pair feature extractor using Transformer's self-attention mechanism vs. State-of-the-art methods was evaluated on Identification accuracy. A Transformer-based sequence pair feature extractor achieved high ECG identification accuracy across multiple databases, including 92.70% and 64.16% on ECGIDDB and PTBDB over a long time separation.
Synapse has enriched 10 related papers. Consider them for comparative context: