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 6, 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

Attention-Based Dynamic Subspace Learners for Medical Image Analysis.

Sukesh Adiga V, Jose Dolz, Herve Lombaert

    IEEE Journal of Biomedical and Health Informatics
    |June 28, 2022
    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

    Generalizable spinal cord multiple sclerosis lesion segmentation across MRI contrasts, protocols, and centers.

    Multiple sclerosis (Houndmills, Basingstoke, England)·2026
    Same author

    Towards reliable WMH segmentation under domain shift: An application study using maximum entropy regularization to improve uncertainty estimation.

    Computers in biology and medicine·2025
    Same author

    Prompt Learning With Bounding Box Constraints for Medical Image Segmentation.

    IEEE transactions on bio-medical engineering·2025
    Same author

    Towards Foundation Models and Few-Shot Parameter-Efficient Fine-Tuning for Volumetric Organ Segmentation.

    Medical image analysis·2025
    Same author

    Neighbor-aware calibration of segmentation networks with penalty-based constraints.

    Medical image analysis·2025
    Same author

    Harmonizing flows: Leveraging normalizing flows for unsupervised and source-free MRI harmonization.

    Medical image analysis·2025

    This study introduces Dynamic Subspace Learners for medical image analysis, dynamically using multiple learners without needing to pre-set their number. This approach enhances image clustering and retrieval by providing visual interpretability through an attention module.

    Area of Science:

    • Medical Image Analysis
    • Computer Vision
    • Machine Learning

    Background:

    • Learning image similarity is crucial for medical image analysis tasks like recommendation systems and anatomical data interpretation.
    • Current methods often use a single metric learner, which is insufficient for capturing diverse image attributes (color, shape, artifacts) and may limit generalization.
    • Existing multi-learner approaches require empirical determination of the number of learners for each dataset.

    Purpose of the Study:

    • To propose a novel method, Dynamic Subspace Learners (DSL), that dynamically utilizes multiple learners for medical image analysis.
    • To eliminate the need for pre-specifying the number of learners, allowing for adaptive aggregation during training.
    • To enhance the visual interpretability of subspace learning by integrating an attention module.

    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

    490

    Related Experiment Videos

    Last Updated: Sep 6, 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

    490

    Main Methods:

    • Developed Dynamic Subspace Learners (DSL) that dynamically exploit multiple learners, aggregating new subspace learners during training.
    • Integrated an attention module to enforce visual interpretability, providing insights into discriminative image features.
    • Evaluated the method's performance in image clustering, image retrieval, and weakly supervised segmentation.

    Main Results:

    • DSL achieved competitive results compared to multi-learner baselines and significantly outperformed a standard classification network on clustering and retrieval tasks across three benchmark datasets.
    • The integrated attention mechanism generated attention maps during inference, offering visual explanations of embedding features.
    • These attention maps acted as proxy labels, improving segmentation accuracy by up to 15% in Dice scores compared to state-of-the-art interpretation techniques.

    Conclusions:

    • Dynamic Subspace Learners offer an effective and interpretable approach for medical image analysis, outperforming existing methods in key tasks.
    • The dynamic aggregation of subspace learners and integrated attention mechanism provide superior performance and visual insights.
    • The method's ability to generate proxy labels via attention maps significantly enhances weakly supervised segmentation tasks.