Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Deep Arm/Ear-ECG Image Learning for Highly Wearable Biometric Human Identification.

Qingxue Zhang1,2,3, Dian Zhou4,5

  • 1Department of Electrical Engineering, University of Texas at Dallas, 800 W Campbell Rd, Richardson, TX, 75080, USA. qingxue.zhg@gmail.com.

Annals of Biomedical Engineering
|October 15, 2017
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Construction of a predictive nomogram for ICU mortality in patients with cardiac arrest using the MIMIC-IV database and the eICU Collaborative Research Database.

BMC cardiovascular disorders·2026
Same author

A proliferation-inducing ligand enhances the anti-tumor effect of BCMA CAR T-cell through inhibition of soluble BCMA.

Biochimica et biophysica acta. Molecular basis of disease·2026
Same author

SlTWD1 Affects Leaf Growth Primarily via Cell Expansion and Contributes to Leaf and Fruit Size Control in Tomato.

Physiologia plantarum·2026
Same author

Late phase transfusion after CAR T therapy is associated with persistent hematotoxicity and non-relapse mortality in multiple myeloma: a post hoc analysis.

BMC medicine·2026
Same author

A multispectral and LC-MS/MS approach for investigating matrix interference in aflatoxin B1 analysis of pressed peanut cake.

Food chemistry·2026
Same author

Anti-BCMA/GPRC5D CAR T in relapsed or refractory multiple myeloma patients with extraosseous extramedullary disease.

Blood·2026

This study introduces a highly wearable electrocardiogram (ECG) system for user identification using non-standard electrode placements. The novel deep learning approach achieves high accuracy, advancing smart health security.

Area of Science:

  • Biomedical Engineering
  • Computer Science
  • Cybersecurity

Background:

  • Increasing security and privacy demands in smart health applications.
  • Limitations of traditional Electrocardiogram (ECG) user identification systems in terms of wearability.
  • Need for convenient and unobtrusive biometric solutions.

Purpose of the Study:

  • To propose a novel, highly wearable ECG-based user identification system.
  • To explore non-standard, convenient ECG lead configurations for enhanced user convenience.
  • To leverage deep learning for accurate identification from weak ECG signals.

Main Methods:

  • Development of a wearable prototype with ECG electrodes on the left upper arm or behind the ears.
  • A two-stage framework involving ECG imaging (projecting heartbeats to a 2D state space) and deep feature learning.
Keywords:
BiometricConvolutional neural networkDeep learningECGMachine learningRepresentation learningSmart healthUser identificationWearable computers

Related Experiment Videos

  • Utilizing a convolutional neural network (CNN) for automatic feature extraction and user identification from ECG images.

    • Main Results:

    • Achieved a promising identification rate of 98.4% using single-arm ECG.
    • Demonstrated a 91.1% identification rate with ear-lead ECG.
    • Successfully obtained distinguishable ECG waveforms from non-standard electrode placements.

    Conclusions:

    • The study demonstrates the feasibility of using single-arm or ear-lead ECG for user identification.
    • The proposed system offers a highly wearable and convenient solution for ECG-based biometrics.
    • This research contributes to the advancement of pervasive user identification in smart health applications.