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 Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A lightweight network for segmenting tree-like structures in medical images.

Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society·2026
Same author

Reconstruction of the Achilles tendon using LARS artificial ligament in bilateral xanthoma: a case report and literature review.

BMC surgery·2026
Same author

Affinity-Tuned Albumin Hitchhiking Extends the Bioorthogonal Capture Window in Pretargeting Radiotheranostics.

Advanced healthcare materials·2026
Same author

The effect of active warming and conventional body temperature management on perioperative blood loss and coagulation function in patients undergoing scoliosis correction: a multicenter randomized controlled trial.

European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society·2026
Same author

Multimodal animal health monitoring in extensive livestock production systems.

Frontiers in veterinary science·2026
Same author

Cell segmentation in microscopy images using a SAM-based U-Net architecture and a novel dataset.

Computer methods and programs in biomedicine·2026

Related Experiment Video

Updated: Sep 11, 2025

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

43.0K

EPDiff: Erasure Perception Diffusion Model for Unsupervised Anomaly Detection in Preoperative Multimodal Images.

Jiazheng Wang, Min Liu, Wenting Shen

    IEEE Transactions on Medical Imaging
    |August 11, 2025
    PubMed
    Summary
    This summary is machine-generated.

    This study introduces the Erasure Perception Diffusion model (EPDiff) for unsupervised anomaly detection in multimodal images. EPDiff effectively reconstructs sub-healthy structures and improves anomaly detection accuracy, outperforming existing methods.

    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

    2.9K
    A Pipeline for 3D Multimodality Image Integration and Computer-assisted Planning in Epilepsy Surgery
    09:41

    A Pipeline for 3D Multimodality Image Integration and Computer-assisted Planning in Epilepsy Surgery

    Published on: May 20, 2016

    12.4K

    Related Experiment Videos

    Last Updated: Sep 11, 2025

    Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
    13:44

    Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

    Published on: August 30, 2013

    43.0K
    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
    A Pipeline for 3D Multimodality Image Integration and Computer-assisted Planning in Epilepsy Surgery
    09:41

    A Pipeline for 3D Multimodality Image Integration and Computer-assisted Planning in Epilepsy Surgery

    Published on: May 20, 2016

    12.4K

    Area of Science:

    • Medical Imaging
    • Artificial Intelligence
    • Computer Vision

    Background:

    • Unsupervised anomaly detection (UAD) struggles with sub-healthy areas mimicking anomalies.
    • Multimodal images offer complementary information to improve UAD.
    • Existing UAD methods often fail to accurately reconstruct complex structures near anomalies.

    Purpose of the Study:

    • To propose a novel unsupervised anomaly detection method for preoperative multimodal images.
    • To enhance the reconstruction of sub-healthy structures around anomalies.
    • To improve the accuracy and robustness of anomaly detection using multimodal data.

    Main Methods:

    • Erasure Perception Diffusion model (EPDiff) incorporating Local Erasure Progressive Training (LEPT) and Global Structural Perception (GSP) modules.
    • LEPT uses a two-phase process to learn sub-healthy structures by progressively erasing and reconstructing anomalies.
    • GSP module captures global structural correlations within and between image modalities.
    • Multimodal Attention Fusion (MAF) module for training-free fusion of anomaly maps.

    Main Results:

    • EPDiff demonstrated superior performance on BraTS2021 and Shifts datasets.
    • Achieved a 2% improvement in AUPRC and 3.9% in mDice on BraTS2021.
    • Achieved a 5.2% improvement in AUPRC and 4.5% in mDice on Shifts.
    • Outperformed state-of-the-art methods in unsupervised anomaly detection tasks.

    Conclusions:

    • EPDiff effectively addresses limitations of traditional UAD methods by leveraging multimodal data.
    • The proposed method shows significant improvements in detecting and reconstructing anomalies in medical images.
    • EPDiff exhibits broad applicability in diverse anomaly diagnosis scenarios.