Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
August 25, 2021ElectronicsOpen Access

A Novel Two-Level Fusion Feature for Mixed ECG Identity Recognition

View Full Paper
Ask AI
Bookmark
Share

Key result

Two-level fusion PCANet achieves ~99.8% recognition accuracy for individual identification.

  • n=426

Why the study?

Training a single ECG database struggles to meet data size and accuracy requirements for biometric identification, creating a need for recognition models capable of processing multi-source data.

Does a two-level fusion PCANet deep recognition network achieve high accuracy for ECG-based biometric identity recognition across mixed databases?

Population

426 individuals from ECG-ID, MIT-BIH, and PTB public databases

Design

Algorithm development and validation study

Authors

XLXin LiuYSYujuan SiWYWeiyi Yang

Discussion

Loading...

Member takes

Overview

Supports ECG biometrics feasibility in mixed databases; leaves open prospective clinical and security validation.

Structured PICO

Does a two-level fusion PCANet deep recognition network achieve high accuracy for ECG-based biometric identity recognition across mixed databases?

P
Population
426 individuals from mixed public databases (ECG-ID, MIT-BIH, and PTB) used to evaluate an ECG identification system.
I
Intervention
Two-level fusion PCANet deep recognition network (combining Hilbert transform, power spectrum, PCANet, MaxFusion algorithm, and linear SVM)
O
Outcome
Recognition accuracy for individual identificationsurrogate

A novel two-level fusion PCANet deep recognition network achieves near-perfect accuracy (99.77%) for biometric identity recognition using mixed ECG databases.

Cite This Study

Liu et al. (2021) studied Biometric identification (n=426). Two-level fusion PCANet deep recognition network was evaluated on Recognition accuracy. A novel two-level fusion PCANet deep recognition network achieved a recognition accuracy of 99.77% on a mixed database of 426 individuals.

synapsesocial.com/papers/6aa93ddfbd00adaa8902cb49https://doi.org/10.3390/electronics10172052
View Full Paper
Ask AI
Bookmark
Share