Jove
Visualize
联系我们
JoVE
x logofacebook logolinkedin logoyoutube logo
关于 JoVE
概览领导团队博客JoVE 帮助中心
作者
出版流程编辑委员会范围与政策同行评审常见问题投稿
图书馆员
用户评价订阅访问资源图书馆顾问委员会常见问题
研究
JoVE JournalMethods CollectionsJoVE Encyclopedia of Experiments存档
教育
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab Manual教师资源中心教师网站
使用条款与条件
隐私政策
政策

相关概念视频

您也可能阅读

相关文章

通过共同作者、期刊和引用图与本文相关的文章。

排序
Same author

Cross-Image Federated Learning for Hyperspectral Image Classification.

IEEE transactions on neural networks and learning systems·2026
Same author

A charge-reversal prodrug activated by tumor-acidity for selective cancer chemotherapy.

Biomaterials science·2026
Same author

ROS-Responsive Cationic Nanoparticles for Cyclosporine A Delivery in Dry Eye Disease: A Dual-Functional Nanotherapeutic Approach.

ACS omega·2026
Same author

PID: A Parameter-Efficient Isolation Domain-Incremental Learning Framework for Signal Modulation Classification.

IEEE transactions on neural networks and learning systems·2025
Same author

Feature refinement and rethinking attention for remote sensing image captioning.

Scientific reports·2025
Same author

Comparing efficacy and safety of low-dose versus standard-dose antiplatelet therapy in stroke patients: a meta-analysis.

Frontiers in pharmacology·2025
Same journal

Non-contact Heart Sound Measurement by Defocused Speckle Imaging.

IEEE journal of biomedical and health informatics·2026
Same journal

TaxEL: Taxonomy-Enhanced Entity Representation Learning for Biomedical Entity Linking.

IEEE journal of biomedical and health informatics·2026
Same journal

Multimodal Feature Prototype Learning for Interpretable and Discriminative Cancer Survival Prediction.

IEEE journal of biomedical and health informatics·2026
Same journal

CrossSG-DTA: Synergizing Sequence Semantics and Graph Structures via Cross-Attention for Drug-Target Affinity Prediction.

IEEE journal of biomedical and health informatics·2026
Same journal

FGCSA-Net: A Novel Framework for Medical Report Generation Via Fine-Grained Feature Preservation and Semantic Alignment.

IEEE journal of biomedical and health informatics·2026
Same journal

Med-SORA: Symptom to Organ Reasoning in Abdomen CT Images.

IEEE journal of biomedical and health informatics·2026
查看所有相关文章

相关实验视频

Updated: Jan 10, 2026

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

DMformer:用于半监督医疗图像分割的难度适应的掩饰变压器.

Zelin Peng, Guanchun Wang, Zhengqin Xu

    IEEE journal of biomedical and health informatics
    |November 26, 2025
    PubMed
    概括
    此摘要是机器生成的。

    这项研究引入了一种新的难度适应罩式变压器 (DMformer),用于半监督的医疗图像细分. DMformer通过调整重建难度来增强学习,显著提高了对有限的标记数据的细分精度.

    更多相关视频

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

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

    Published on: July 5, 2024

    727
    Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
    06:48

    Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images

    Published on: January 7, 2019

    9.4K

    相关实验视频

    Last Updated: Jan 10, 2026

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

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

    Published on: July 5, 2024

    727
    Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
    06:48

    Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images

    Published on: January 7, 2019

    9.4K

    科学领域:

    • 医学图像分析 医学图像分析
    • 计算机视觉 计算机视觉
    • 机器学习 机器学习

    背景情况:

    • 利用未标记的数据对于半监督的医疗图像细分至关重要.
    • 共享人体解剖学为利用未标记的医学图像提供了强有力的先验.
    • 蒙面图像建模激发了新的方法,用于结合解剖学先验.

    研究的目的:

    • 开发一个半监督的医疗图像细分框架,有效地利用通过解剖学先验的未标记数据.
    • 引入适应难度的掩护机制,以处理不同器官/组织的不同复杂性重建.
    • 为了提高医疗图像细分模型的性能,使用有限的标记数据.

    主要方法:

    • 将辅助无监督的粗体解剖学重建任务纳入了教师与学生的框架.
    • 通过调节蒙面区域和类比,开发了一个适应难度的面具机制.
    • 实施基于区域和基于阶级的掩盖策略,根据重建难度量身定制.
    • 使用了冲突意识梯度计算策略来管理同时因子调制.
    • 在视觉变压器上构建了适应困难的蒙面变压器 (DMformer).

    主要成果:

    • 在半监督的医疗图像细分中,DMformer表现出卓越的性能.
    • 在ACDC和Synapse数据集上的子相似系数 (DSC) 中取得了显著的改进.
    • 超越了最先进的 (SOTA) 方法,在ACDC上有5%的标记图像 (9.53%的DSC改进).
    • 超越了SOTA方法,在Synapse上有30%的标记图像 (4.63%的DSC改进).

    结论:

    • 拟议的难度适应的面具机制有效地解决了医疗图像细分中的各种重建挑战.
    • DMformer为半监督医疗图像细分提供了一种强大的方法,特别是在有限的标记数据下.
    • 该框架通过辅助重建任务成功地利用解剖学先验,提高了细分的准确性.