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 Video

Updated: May 26, 2026

Soft Pneumatic Robot Modulates Graph Theory Metrics of Brain Network for Hand Rehabilitation After Stroke
05:30

Soft Pneumatic Robot Modulates Graph Theory Metrics of Brain Network for Hand Rehabilitation After Stroke

Published on: October 10, 2025

Spatio-temporal graph attention network for rehabilitation movement classification.

Qiyu Yang1, Zehui Zhang1, Yi Huang2

  • 1School of Automation, Guangdong University of Technology, Guangzhou, China.

Medical & Biological Engineering & Computing
|May 25, 2026
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

Electrical impedance spectroscopy of young and old mouse multiple tissues.

Biomedical physics & engineering express·2026
Same author

Bridging aging and colorectal cancer: synergistic roles of inflammaging and immunosenescence.

Frontiers in immunology·2026
Same author

Metaphase I-arrest oocyte with agar-like zona pellucida, abandon or not abandon?

Journal of ovarian research·2026
Same author

The ELD4-OsPRR95 module represses OsMADS51 in regulating rice heading date.

The Plant cell·2026
Same author

Machine learning-driven renewable energy grid integration stability assessment: LIME interpretability and LLM intelligent analysis.

Scientific reports·2026
Same author

Impact of post-vitrification storage duration on clinical outcomes of frozen-thawed cleavage-stage embryos transfer: a propensity score-matched retrospective cohort study.

Frontiers in endocrinology·2026

This study introduces STGA-Net, a novel framework for automated physical rehabilitation assessment. It accurately monitors patient movements using skeleton data, enhancing home-based and clinical recovery.

Area of Science:

  • Biomedical Engineering
  • Computer Science
  • Rehabilitation Medicine

Background:

  • Skeleton-based methods offer privacy-preserving analysis of patient movements for home-based rehabilitation.
  • Existing methods struggle to capture the full temporal dynamics of rehabilitation exercises.
  • A need exists for robust, automated tools for quantitative rehabilitation assessment.

Purpose of the Study:

  • To develop a novel cascaded framework, STGA-Net, for comprehensive assessment of home-based rehabilitation.
  • To improve the modeling of temporal coherence in exercise sequences using skeleton data.
  • To provide a quantitative and privacy-preserving tool for personalized rehabilitation.

Main Methods:

  • A cascaded framework combining a Spatio-Temporal Graph Convolutional Network (ST-GCN) and a Transformer-based attention mechanism.
Keywords:
Home-based RehabilitationHuman Activity RecognitionPhysical RehabilitationSkeleton Data

Related Experiment Videos

Last Updated: May 26, 2026

Soft Pneumatic Robot Modulates Graph Theory Metrics of Brain Network for Hand Rehabilitation After Stroke
05:30

Soft Pneumatic Robot Modulates Graph Theory Metrics of Brain Network for Hand Rehabilitation After Stroke

Published on: October 10, 2025

  • ST-GCN captures local joint dynamics, while the Transformer models global, long-range dependencies.
  • Rigorous evaluation using 5-fold cross-validation on the UI-PRMD and KIMORE datasets.

    • Main Results:

    • STGA-Net achieved high mean accuracies of 92.71% on the UI-PRMD dataset and 95.24% on the KIMORE dataset.
    • The proposed method significantly surpasses existing state-of-the-art approaches in rehabilitation assessment.
    • Demonstrated effectiveness in capturing both local and global movement patterns.

    Conclusions:

    • STGA-Net represents a significant advancement in automated rehabilitation assessment.
    • The framework offers a quantitative and privacy-preserving solution for personalized rehabilitation.
    • Potential for widespread application in both clinical settings and home-based recovery monitoring.