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

The impact of vision care program on the mental health of migrant children in Eastern China: evidence from a cluster-randomized controlled trial.

International journal for equity in health·2026
Same author

Nanoplastics exposure accelerates the progression of osteoarthritis via lysosomal destabilization-mediated pyroptosis.

Journal of advanced research·2026
Same author

Global trends and projections of lung cancer attributable to residential radon exposure, 1990-2050.

BMC cancer·2026
Same author

Data-Driven Material Selection for Flexible Wearable Sensors Under Environmental Coupling Conditions.

Sensors (Basel, Switzerland)·2026
Same author

RIPK1 ubiquitination regulates its kinase-independent function in development and inflammation.

Proceedings of the National Academy of Sciences of the United States of America·2026
Same author

Biomimetic self-assembled microspheres based on selenized Angelica-Astragalus compound polysaccharides and platelet membranes for targeted therapy of acute liver injury.

Biomedical materials (Bristol, England)·2026

相关实验视频

Updated: Dec 30, 2025

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

3.6K

非IID医疗图像细分基于各种多中心场景的级联扩散模型.

Hanwen Zhang, Mingzhi Chen, Yuxi Liu

    IEEE journal of biomedical and health informatics
    |March 27, 2025
    PubMed
    概括

    本研究引入了一个保护隐私的框架,用于对非独立且相同分布的 (非IID) 医疗数据进行联合学习. 该方法增强了医疗图像细分模型,提高了准确性和降低了通信成本.

    科学领域:

    • 人工智能的人工智能
    • 医疗成像医学成像
    • 机器学习 机器学习

    背景情况:

    • 联合学习 (FL) 在多中心医疗数据集中面临挑战,原因是数据异质性和隐私问题.
    • 现有的FL方法与非独立且相同分布 (非IID) 数据扎,并产生高通信成本.

    研究的目的:

    • 提出一个实用的隐私保护框架,用于培训非IID医疗图像细分模型在多中心设置中,通信开销较低.
    • 减轻数据异质性,提高医疗保健中联合学习的效率.

    主要方法:

    • 开发了一种高效的级联扩散模型来生成合成图像-面具对,解决数据异质性.
    • 实施了标签构建模块,以提高生成数据的质量.
    • 建议的聚合方法 (CD-Syn,CD-Ens,CD-KD) 适用于不同的场景,平衡效率,隐私和准确性.

    主要成果:

    • 与基线FednnU-Net相比,该框架在五个非IID医疗数据集中实现了平均5.38%的Dice得分改善.
    • 在实用的多中心环境中证明有效性,对隐私和准确性的要求各不相同.
    • 成功地降低了通信成本,同时保持了高性能.

    结论:

    更多相关视频

    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
    Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
    10:25

    Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

    Published on: September 25, 2019

    49.2K

    相关实验视频

    Last Updated: Dec 30, 2025

    Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
    04:25

    Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

    Published on: December 15, 2023

    3.6K
    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
    Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
    10:25

    Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping

    Published on: September 25, 2019

    49.2K
    • 拟议的隐私保护框架为使用联合学习进行非IID医疗图像细分提供了灵活和有效的解决方案.
    • 级联扩散模型和新的聚合策略显著提高了模型性能和数据实用性.
    • 这种方法为安全和高效地利用多中心医疗数据提供了一个实用的途径.