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: Mar 6, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

1.2K

Self-Supervised Endoscopic Depth Estimation via Deep Feature-Aware Reconstruction and Dual-Path Feature Aggregation

Yukang Ren1, Yanping Chen2

  • 1School of Mechanical and Electrical Engineering, Chengdu University of Technology, Chengdu, China. renyukang@stu.cdut.edu.cn.

Journal of Imaging Informatics in Medicine
|March 4, 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

Efficacy and safety of inhaled ambroxol hydrochloride solution in Chinese pediatric patients with acute lower respiratory tract infections: a real-world, multicenter, open-label, single-arm study.

Translational pediatrics·2026
Same author

Upcycling lignocellulosic biomass using a radial-axial fixed-bed ex-situ catalytic pyrolysis reactor.

Bioresource technology·2026
Same author

Traditional Chinese Medicine Nursing Intervention in Chronic Obstructive Pulmonary Disease With Gastrointestinal Dysfunction: Bibliometric and Knowledge Graph Analysis, 2015-2025.

Complementary medicine research·2026
Same author

Proteomic clocks combined with deep learning phenotypes track eye aging and diseases.

NPJ digital medicine·2026
Same author

Folding integrated all-paper photoelectrochemical immunoassay using annealed ZnO for point-of-care detection of ferritin.

Analytica chimica acta·2026
Same author

Chemical residues in declining commercial honey bee colonies.

Environmental toxicology and pharmacology·2026

This study introduces a new AI framework for endoscopic depth estimation, improving 3D spatial guidance in minimally invasive surgeries. The innovative Dual-Path Feature Aggregation Pyramid Module and Deep Feature-Aware Reconstruction Module significantly enhance accuracy and generalization.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computer Vision

Background:

  • Endoscopes are crucial for minimally invasive procedures in gastrointestinal, respiratory, and urinary tracts.
  • Accurate depth estimation in endoscopic imaging is vital for enhancing surgical precision and patient safety.
  • Existing methods face limitations in hierarchical information interaction and semantic consistency.

Purpose of the Study:

  • To develop an advanced AI framework for endoscopic depth estimation.
  • To improve 3D spatial guidance during minimally invasive surgeries.
  • To overcome limitations of traditional depth estimation techniques.

Main Methods:

  • Proposed a novel framework with a Dual-Path Feature Aggregation Pyramid Module (DPFAP) for hierarchical alignment and fusion.
Keywords:
Deep feature reconstructionEndoscopic depth estimationFeature fusionPhotometric lossSelf-supervised learning

More Related Videos

Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility
07:46

Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility

Published on: August 9, 2024

1.2K

Related Experiment Videos

Last Updated: Mar 6, 2026

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

1.2K
Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility
07:46

Author Spotlight: Revolutionizing Remote Surgery with Augmented Reality and Robotics for Enhanced Precision and Accessibility

Published on: August 9, 2024

1.2K
  • Introduced a Deep Feature-Aware Reconstruction Module (DFAR) with a feature reconstruction loss strategy for unsupervised learning.
  • Utilized the SCARED and Hamlyn endoscopic datasets for training and validation.

    • Main Results:

    • Achieved state-of-the-art performance on the SCARED dataset with significant improvements in Abs Rel, RMSE, and delta metrics.
    • Demonstrated a 3.9% reduction in Abs Rel and a 1.9% reduction in RMSE compared to existing methods.
    • Showcased excellent generalization capability and scene adaptability on the Hamlyn dataset.

    Conclusions:

    • The proposed DPFAP and DFAR modules offer a robust and effective solution for endoscopic depth estimation.
    • The framework provides superior 3D spatial guidance, enhancing safety in minimally invasive procedures.
    • The algorithm's strong performance and generalization validate its practical utility in diverse endoscopic applications.