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: Jan 7, 2026

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
08:20

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

2.4K

Hybrid Dual-Heterogeneous Knowledge Distillation Network for Anomaly Detection in Retinal OCT Images.

Muhao Xu, Hua Wei, Zihan Nie

    IEEE Journal of Biomedical and Health Informatics
    |January 2, 2026
    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

    Effect of iTBS on suicidal ideation and emotion regulation in major depressive disorder: a randomized clinical trial.

    BMC psychiatry·2026
    Same author

    Baseline Anhedonia Phenotypes Reveal Divergent Neurophysiological Effects of Intermittent Theta-Burst Stimulation in Major Depression Disorder.

    Biological psychiatry. Cognitive neuroscience and neuroimaging·2026
    Same author

    Risk prediction model of survival in patients with low-grade serous ovarian cancer: a multicenter Cohort study.

    Journal of ovarian research·2026
    Same author

    Targeting magnesium homeostasis: a novel therapeutic strategy for liver diseases.

    Frontiers in nutrition·2026
    Same author

    Prefrontal-limbic dysconnectivity tracks emotion regulation deficits in depressed adolescents with social anxiety symptom severity.

    Journal of affective disorders·2026
    Same author

    Comparative effects of intermittent theta-burst stimulation and sequential bilateral rTMS on depression and emotion regulation in major depressive disorder: a randomized controlled trial.

    Frontiers in neuroscience·2026
    Same journal

    Graph Convolutional Neural Network based Depression Detection using Brain Functional Connectivity Measures.

    IEEE journal of biomedical and health informatics·2026
    Same journal

    Improving Multi-Sensor Non-Invasive Glucose Detection through AI: A Domain Generalization Approach.

    IEEE journal of biomedical and health informatics·2026
    Same journal

    Unmixing the Neck: Accurate Jugular Venous Pulse Detection From Wearable PPG.

    IEEE journal of biomedical and health informatics·2026
    Same journal

    AD-DAE: Alzheimer's Disease Progression Modeling with Unpaired Longitudinal MRI using Diffusion Auto-Encoders.

    IEEE journal of biomedical and health informatics·2026
    Same journal

    EEG Connectivity Signatures in Active vs. Passive Mental Fatigue Settings.

    IEEE journal of biomedical and health informatics·2026
    Same journal

    Privacy-Enhanced Vertical Federated Learning for Healthcare via Directional Noise and Subset Representations.

    IEEE journal of biomedical and health informatics·2026
    See all related articles

    This study introduces a new hybrid dual-heterogeneous knowledge distillation network for unsupervised medical anomaly detection in retinal Optical Coherence Tomography (OCT) images. The method improves detection of diverse anomalies by using distinct teacher and student network architectures.

    Area of Science:

    • Medical Imaging
    • Computer Vision
    • Artificial Intelligence

    Background:

    • Unsupervised medical anomaly detection identifies abnormalities using only normal training data, crucial for rare diseases.
    • Knowledge distillation methods compare teacher and student networks but suffer from identity mapping issues due to similar architectures.
    • Retinal Optical Coherence Tomography (OCT) images present challenges in anomaly detection due to diverse lesion types.

    Purpose of the Study:

    • To propose a novel hybrid dual-heterogeneous knowledge distillation network for unsupervised anomaly detection in retinal OCT images.
    • To address the identity mapping problem and enhance sensitivity to both structural and logical anomalies.
    • To achieve state-of-the-art performance in detecting diverse anomaly types in medical images.

    Main 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

    3.3K

    Related Experiment Videos

    Last Updated: Jan 7, 2026

    Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
    08:20

    Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

    Published on: October 27, 2023

    2.4K
    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

    3.3K
    • A hybrid dual-heterogeneous knowledge distillation network with an encoder-only teacher and an encoder-dual decoder student architecture.
    • Multi Feature Model using convolutional and depthwise convolutional blocks for local feature extraction and structural anomaly detection.
    • Mamba UpNet employing self-supervised learning for long-range dependency capture and global anomaly pattern identification.

    Main Results:

    • The proposed heterogeneous network design effectively mitigates the identity mapping problem.
    • The method demonstrates enhanced sensitivity to both structural and logical anomalies.
    • Achieved state-of-the-art performance on two retinal OCT anomaly detection datasets, handling diverse anomaly types.

    Conclusions:

    • The novel hybrid dual-heterogeneous knowledge distillation network offers a superior approach for unsupervised medical anomaly detection.
    • The distinct network architectures and feature extraction methods improve the detection of complex anomalies in retinal OCT images.
    • This work provides a robust solution for identifying disease-related irregularities without requiring extensive labeled datasets.