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: Sep 14, 2025

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

2.9K

MAFL-Attack: a targeted attack method against deep learning-based medical image segmentation models.

Junmei Sun1, Xin Zhang1, Xiumei Li1

  • 1Hangzhou Normal University, School of Information Science and Technology, Hangzhou, China.

Journal of Medical Imaging (Bellingham, Wash.)
|July 21, 2025
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

Uncertainty-aware collaborative learning with mixed images for semi-supervised medical image segmentation.

Computer methods and programs in biomedicine·2025
Same author

DECE-Net: a dual-path encoder network with contour enhancement for pneumonia lesion segmentation.

Journal of medical imaging (Bellingham, Wash.)·2025
Same author

OsOFP9 regulates diverse key traits of rice by integrating multiple plant hormones.

The Plant journal : for cell and molecular biology·2025
Same author

Effects of Longitudinal External Magnetic Field on Metal Transfer Behavior and Spatter Formation in CO<sub>2</sub> Arc Welding.

Materials (Basel, Switzerland)·2025
Same author

Research Progress on the Modification of B and Al High-Energy Fuels for Powder Fuel Ramjet Applications.

ACS applied materials & interfaces·2025
Same author

An interpretable machine learning model assists in predicting induction chemotherapy response and survival for locoregionally advanced nasopharyngeal carcinoma using MRI: a multicenter study.

European radiology·2025

Researchers developed MAFL-Attack to improve adversarial attacks on deep learning medical image segmentation. This method enhances attack effectiveness and image quality, strengthening model robustness against misdiagnosis.

Area of Science:

  • Artificial Intelligence
  • Medical Imaging
  • Computer Vision

Background:

  • Deep learning models are vital for computer-aided medical diagnosis via image segmentation.
  • These models are vulnerable to adversarial attacks, potentially causing misdiagnosis.
  • Existing adversarial attacks on medical image segmentation lack effectiveness and image quality.

Purpose of the Study:

  • To propose a novel adversarial attack method for deep learning-based medical image segmentation.
  • To address limitations of existing methods, focusing on targeted attacks and improved adversarial example quality.
  • To enhance the robustness of medical image segmentation models against sophisticated attacks.

Main Methods:

  • Introduced MAFL-Attack (momentum-driven adaptive feature-cosine-similarity with low-frequency constraint attack).
Keywords:
adversarial exampledeep learninglow-frequency component constraintmedical image segmentationtargeted attack

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

527

Related Experiment Videos

Last Updated: Sep 14, 2025

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

2.9K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

527
  • Employed feature-cosine-similarity loss to disrupt model understanding of adversarial examples.
  • Utilized low-frequency component constraint for imperceptibility and momentum with dynamic step-size for attack enhancement.

    • Main Results:

    • MAFL-Attack demonstrated superior targeted attack effects compared to Adaptive Segmentation Mask Attack.
    • Achieved better performance across metrics including Intersection over Union, accuracy, L2, L-infinity, PSNR, and SSIM.
    • Generated adversarial examples with improved attack efficacy and image quality.

    Conclusions:

    • The MAFL-Attack method offers a significant advancement in adversarial attack research for medical image segmentation.
    • Findings highlight the vulnerability of current models and the need for robust defenses.
    • The proposed approach inspires the development of defensive strategies to strengthen model resilience.