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: Nov 17, 2025

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

650

Segmentation-Renormalized Deep Feature Modulation for Unpaired Image Harmonization.

Mengwei Ren, Neel Dey, James Fishbaugh

    IEEE Transactions on Medical Imaging
    |February 16, 2021
    PubMed
    Summary
    This summary is machine-generated.

    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

    MultiMorph: On-demand Atlas Construction.

    Proceedings. IEEE Computer Society Conference on Computer Vision and Pattern Recognition·2026
    Same author

    Treatment Outcomes With Novel Targeted and Immunotherapeutic Regimens in CAYA Hodgkin Lymphoma: A Retrospective Study.

    Cancer innovation·2026
    Same author

    Consistency Regularization Improves Placenta Segmentation in Fetal EPI MRI Time Series.

    Perinatal, preterm and paediatric image analysis (2025)·2026
    Same author

    Learning General-Purpose Biomedical Volume Representations using Randomized Synthesis.

    ... International Conference on Learning Representations·2026
    Same author

    Robust Fetal Pose Estimation across Gestational Ages via Cross-Population Augmentation.

    Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention·2026
    Same author

    New insights into the treatment of nasopharyngeal carcinoma in children, adolescents, and young adults: a retrospective study.

    Frontiers in immunology·2026

    This study introduces a novel image harmonization method for multi-center medical imaging studies. The segmentation-renormalization framework improves image quality and downstream analysis accuracy across different scanners and modalities.

    Area of Science:

    • Medical Imaging
    • Artificial Intelligence
    • Computer Vision

    Background:

    • Deep networks are widely used in multi-center imaging studies.
    • Image aggregation across sites is challenging due to variations in contrast, resolution, and noise.
    • Existing Cycle-consistent Generative Adversarial Networks (GANs) for harmonization have limitations like instability and pathology manipulation.

    Purpose of the Study:

    • To develop a robust image harmonization framework for multi-center medical imaging.
    • To address the limitations of existing GAN-based harmonization methods.
    • To preserve anatomical layout while reducing inter-scanner heterogeneity.

    Main Methods:

    • Proposed a segmentation-renormalized image translation framework.
    • Replaced affine transformations with trainable scale and shift parameters conditioned on segmentation embeddings.

    More Related Videos

    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

    3.1K

    Related Experiment Videos

    Last Updated: Nov 17, 2025

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

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

    Published on: July 5, 2024

    650
    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

    3.1K
  • Modulated features at every level of translation within generative networks.

    • Main Results:

    • Achieved superior image harmonization compared to recent baselines across T1w MRI, FLAIR MRI, and OCT modalities.
    • Demonstrated improved downstream utility through enhanced post-hoc segmentation accuracy.
    • Showcased improved robustness against translation perturbation and self-adversarial attacks.

    Conclusions:

    • Segmentation-renormalization offers a stable and reliable approach for medical image harmonization.
    • The proposed method effectively reduces inter-scanner heterogeneity while preserving crucial anatomical information.
    • This framework has the potential for reliable adoption in real-world medical imaging analysis.