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

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

3.6K

MDRSGAN : Multi-Scale Deep Residual Shrinkage Generative Adversarial Network for Medical Image Enhancement.

Du Jiang, Bo Tao, Kezhi Deng

    IEEE Journal of Biomedical and Health Informatics
    |March 4, 2026
    PubMed
    Summary

    This study introduces a new AI network, MDRSGAN, for enhancing medical images. The Multi-scale Deep Residual Shrinkage Generative Adversarial Network (MDRSGAN) improves image quality for better diagnosis.

    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

    WITHDRAWN: Reassessing the link between PA and CKD in CMD patients.

    Journal of science and medicine in sport·2026
    Same author

    Associations between visual skills and serve reception performance in elite male volleyball outside hitters.

    Frontiers in physiology·2026
    Same author

    Terahertz Dielectric Characterization and Hybrid Debye-Lorentz Modeling of Silicone Rubber Composites for Composite Insulators.

    Polymers·2026
    Same author

    The effects of stroboscopic vision training on sports vision ability performance and punching performance in female boxers.

    Frontiers in physiology·2026
    Same author

    Excitotoxicity and PANoptosis are examined through endoplasmic reticulum stress of brain in the fat greenling (Hexagrammos otakii) under hypoxic stress.

    Fish & shellfish immunology·2026
    Same author

    Lightweight detection network based on bidirectional weighted feature fusion with small target enhancement for USVs.

    Scientific reports·2026

    Area of Science:

    • Medical Imaging
    • Artificial Intelligence
    • Computer Vision

    Background:

    • High-quality medical images are crucial for accurate clinical diagnosis and treatment.
    • Image degradation from noise, artifacts, and uneven lighting poses a significant challenge.
    • Existing enhancement methods often struggle with complex image quality issues.

    Purpose of the Study:

    • To propose a novel Multi-scale Deep Residual Shrinkage Generative Adversarial Network (MDRSGAN) for unpaired medical image enhancement.
    • To address limitations in current medical image quality improvement techniques.
    • To develop a robust method for enhancing diverse medical imaging modalities.

    Main Methods:

    • Developed a customized generator (DRSN-CS) for hierarchical feature extraction and adaptive noise suppression.

    More Related Videos

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
    04:48

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

    Published on: July 5, 2024

    851

    Related Experiment Videos

    Last Updated: Mar 6, 2026

    Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
    04:48

    Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

    Published on: November 30, 2022

    3.6K
    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
    04:48

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

    Published on: July 5, 2024

    851
  • Employed a dual-core discriminator to ensure global statistical consistency and local structural fidelity.
  • Integrated content perception loss and lighting loss for optimizing image details and features.

    • Main Results:

    • MDRSGAN demonstrated superior performance on fundus retina, endoscope, and NIR-II mouse datasets compared to five mainstream methods.
    • Achieved a 5% increase in Signal-to-Noise Ratio (SNR) for NIR-II mouse images.
    • Improved Intersection over Union (IoU) and Dice Similarity Coefficient (DSC) by 13% and 8% in downstream retinal vessel segmentation.

    Conclusions:

    • The proposed MDRSGAN effectively enhances medical image quality, outperforming existing methods.
    • The network's architecture and loss functions contribute to significant improvements in image fidelity and detail.
    • Demonstrated strong clinical applicability and performance advantages, particularly in enhancing retinal images for segmentation tasks.