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

Calculation of state-to-state differential and integral cross sections for atom-diatom reactions with transition-state wave packets.

The Journal of chemical physics·2014
Same author

[Natural attenuation of tetracycline in the water of Taihu Lake under different environmental conditions].

Huan jing ke xue= Huanjing kexue·2014
Same author

The role of AhR in autoimmune regulation and its potential as a therapeutic target against CD4 T cell mediated inflammatory disorder.

International journal of molecular sciences·2014
Same author

Association between polymorphisms in the flanking region of the TAFI gene and atherosclerotic cerebral infarction in a Chinese population.

Lipids in health and disease·2014
Same author

An Updated Analysis with 85,939 Samples Confirms the Association Between CR1 rs6656401 Polymorphism and Alzheimer's Disease.

Molecular neurobiology·2014
Same author

Immobilized lipase from Candida sp. 99-125 on hydrophobic silicate: characterization and applications.

Applied biochemistry and biotechnology·2014

相关实验视频

Updated: Jun 12, 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

2.7K

交叉匹配:通过扰乱策略和知识蒸来增强半监督的医疗图像细分.

Bin Zhao, Chunshi Wang, Shuxue Ding

    IEEE journal of biomedical and health informatics
    |September 18, 2024
    PubMed
    概括

    通过使用有限的标记数据与丰富的未标记数据,CrossMatch增强了医疗图像细分. 这种新的框架通过双扰和知识蒸来提高模型的准确性和稳定性.

    科学领域:

    • 医疗成像医学成像
    • 计算机视觉 计算机视觉
    • 机器学习 机器学习

    背景情况:

    • 由于有限的标记数据,半监督学习对于医学图像细分至关重要.
    • 现有的方法很难充分利用未标记的数据来提高模型性能.
    • 提高细分任务中的模型稳定性和准确性仍然是一个关键的挑战.

    研究的目的:

    • 引入CrossMatch,这是一个用于半监督医疗图像细分的新框架.
    • 改进对标记和未标记数据的利用,以增强模型学习.
    • 提高医疗图像细分模型的稳定性和准确性.

    主要方法:

    • 交叉匹配将知识蒸与双重扰动策略 (图像级和特征级) 整合在一起.
    • 多个编码器和解码器生成各种数据流,用于自知蒸.
    • 这种方法在各种扰动中提高了预测的一致性和可靠性.

    主要成果:

    • 在标准基准指标中,CrossMatch显著超过了最先进的技术.
    • 该方法有效地减少了标记和未标记数据训练之间的性能差距.
    • 在医学图像细分中观察到更好的边缘精度和概括能力.

    更多相关视频

    From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
    12:08

    From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

    Published on: August 13, 2014

    24.5K
    Author Spotlight: Segmentation and VR for Advanced Neurovascular Interventions
    06:18

    Author Spotlight: Segmentation and VR for Advanced Neurovascular Interventions

    Published on: April 5, 2024

    979

    相关实验视频

    Last Updated: Jun 12, 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

    2.7K
    From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
    12:08

    From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

    Published on: August 13, 2014

    24.5K
    Author Spotlight: Segmentation and VR for Advanced Neurovascular Interventions
    06:18

    Author Spotlight: Segmentation and VR for Advanced Neurovascular Interventions

    Published on: April 5, 2024

    979

    结论:

    • 交叉匹配为半监督的医疗图像细分提供了强大的解决方案.
    • 该框架在不增加计算成本的情况下实现了显著的性能改进.
    • 这项研究表明了整合知识蒸和双扰动的有效性.