Intersubject variability and intrasubject reproducibility of 12-lead ECG metrics: Implications for human verification
详细信息 查看全文
- 作者：Irena Jekova ; PhDa ; irena@biomed.bas.bg" class="auth_mail" title="E-mail the corresponding author ; Vessela Krasteva ; PhDa ; Remo Leber ; MSEb ; Ramun Schmid ; MSEb ; Raphael Twerenbold ; MDc ; Christian Mü ; ller ; MDc ; Tobias Reichlin ; MDc ; Roger Abä ; cherli ; PhDb ; c
- 关键词：ECG biometrics ; Person authentication ; ECG diagnostic morphological features ; Optimization criteria ; Linear discriminant analysis
- 刊名：Journal of Electrocardiology
- 出版年：2016
- 出版时间：November-December 2016
- 年：2016
- 卷：49
- 期：6
- 页码：784-789
- 全文大小：869 K
文摘
- •
Optimization of features for high reliability and security of ECG biometrics is done.
- •
12-lead ECG from 574 non-cardiac patients with 2 remote recordings (>1 year) are used.
- •
202 lead-specific features (durations, amplitudes, axes, and ST) are measured.
- •
Optimal set is scored: (intersubject variability, intrasubject reproducibility) → max
- •
Discriminant analysis validates 11 features for person verification (accuracy > 85%)