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

B-EPIC: A Transformer-Based Language Model for Decoding B Cell Immunodominance Patterns.

Advanced science (Weinheim, Baden-Wurttemberg, Germany)·2025
Same author

Ginsenoside Rk2 Provides Multidimensional Relief for Ulcerative Colitis by Improving Intestinal Barrier Function, Reshaping Gut Microbiota, and Regulating Th17/Treg Balance.

Journal of agricultural and food chemistry·2025
Same author

Retraction notice to 'Distribution and potential sources of iodine in particulate matter at an industrial city in Northwest China' [Environ. Res. 274 (2025) 121364].

Environmental research·2025
Same author

Norepinephrine or Phenylephrine to Prevent Spinal Anesthesia-Induced Hypotension During Cesarean Section: A Double-Blinded, Randomized, Controlled Study of Fetal Cerebral Perfusion.

Drug design, development and therapy·2025
Same author

Hierarchical Structures of Protein-Fiber-Infused Aminoglycan Hydrogels to Promote Myocardial Repair.

ACS nano·2025
Same author

Global, regional, and national burden of peripheral artery disease: a systematic analysis of prevalence, incidence, deaths, and DALYs with projections for the next 15 years.

Nutrition, metabolism, and cardiovascular diseases : NMCD·2025

相关实验视频

Updated: Jun 23, 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

无监督的域名适应多层次的基于对抗学习的交叉域名视网膜血管细分.

Jinping Liu1, Junqi Zhao1, Jingri Xiao1

  • 1College of Information Science and Engineering, Hunan Normal University, Changsha, Hunan, 410081, China.

Computers in biology and medicine
|June 25, 2024
PubMed
概括

这项研究引入了一个新的框架,用于在不同的图像来源中准确地对视网膜血管进行细分. 基于多层次对抗性学习和伪标签否定的自我训练框架 (MLAL&PDSF) 显著提高了疾病诊断的跨领域细分精度.

关键词:
多层次的对抗式学习.伪标签拒绝使用.视网膜血管细分 视网膜血管细分无监督的域名适应

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

386
Measuring Retinal Vessel Diameter from Mouse Fluorescent Angiography Images
04:04

Measuring Retinal Vessel Diameter from Mouse Fluorescent Angiography Images

Published on: May 19, 2023

609

相关实验视频

Last Updated: Jun 23, 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
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

386
Measuring Retinal Vessel Diameter from Mouse Fluorescent Angiography Images
04:04

Measuring Retinal Vessel Diameter from Mouse Fluorescent Angiography Images

Published on: May 19, 2023

609

科学领域:

  • 医疗成像医学成像
  • 计算机视觉 计算机视觉
  • 生物医学工程 生物医学工程

背景情况:

  • 视网膜血管细分对于诊断诸如玻璃眼和视网膜病变等疾病至关重要.
  • 目前的模型在与跨源基金图像作斗争,限制了诊断准确度.
  • 准确的细分对于理解和诊断各种系统性和眼部疾病至关重要.

研究的目的:

  • 开发一个强大的框架,在多种 fundus 图像来源中准确地对视网膜血管进行细分.
  • 克服现有模型在处理跨领域细分挑战方面的局限性.
  • 为了提高视网膜血管细分的精度,以改善疾病诊断.

主要方法:

  • 提出了一种新的基于多层次对抗性学习和伪标签拒绝的自我训练框架 (MLAL&PDSF).
  • 在特征和图像层使用多层次的对抗网络来调整源和目标域分布.
  • 利用距离比较来改进伪标签和纠正自我训练中的不准确性.

主要成果:

  • 在多个数据集 (CHASEDB1,STARE,HRF) 上实现了非凡的无监督域自适应细分性能.
  • 在跨域细分任务中 (例如,驱动到CHASEDB1和STARE) 证明了高AUC,灵敏度,特异性,准确性和F1分数.
  • 通过广泛的比较实验验证实该框架的有效性.

结论:

  • MLAL&PDSF有效地实现了跨域视网膜血管数据集的准确细分.
  • 该框架为推进跨领域细分技术提供了坚实的基础.
  • 增强的细分精度有助于更好地诊断和理解视网膜疾病.