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

Classification of Epithelial Tissues: Overview01:22

Classification of Epithelial Tissues: Overview

19.6K
Epithelial tissues are classified according to the shape of the cells and the number of cell layers formed. Cell shapes can be squamous (flattened and thin), cuboidal (square-like, as wide as it is tall), or columnar (rectangular, taller than it is wide). Additionally, the nucleus shape helps identify the type of epithelial cells. Squamous cells have flattened disc-shaped nuclei, cuboidal cells have spherical nuclei, and columnar cells have elongated nuclei.
Based on the number of cell layers,...
19.6K

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

Interpretable machine learning model for predicting kidney failure among CAKUT children in multicenter large-scale study.

NPJ digital medicine·2026
Same author

Modulating the coordination environment of electrochemical catalysis for enhanced catalytic performance.

Chemical communications (Cambridge, England)·2026
Same author

Engineering an Ultrahigh-Surface-Area Diatomic Catalyst via Two-Dimensional-Templated Vapor-Deposition for Advanced Energy Conversion.

Small (Weinheim an der Bergstrasse, Germany)·2026
Same author

Preoperative prediction of lymphatic metastasis in rectal cancer using a fusion model based on multiparameter magnetic resonance imaging: a retrospective validation study.

Frontiers in oncology·2026
Same author

HeyJay! A corpus of atypical speech for spoken language understanding and automatic speech recognition.

Scientific data·2026
Same author

Aggregation-induced emission luminogen in ternary organic bulk-heterojunction for efficient perovskite-organic tandem solar cells.

Nature communications·2026

相关实验视频

Updated: Jan 18, 2026

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
05:33

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System

Published on: July 11, 2025

809

通过基于原型的多个实例学习,对异质基因病理学的结构意识概括.

Zhenjun Yu1, Zhelin Xia1, Donghao Xu2,3

  • 1Taizhou Central Hospital (Taizhou University Hospital), Taizhou, Zhejiang, China.

NPJ digital medicine
|January 15, 2026
PubMed
概括
此摘要是机器生成的。

通过整合结构和原型,以提高准确性和可解释性,StructMIL可以从整个幻灯片图像中加强癌症诊断. 这种计算病理学框架在不同机构的乳腺和前列腺癌分级方面取得了最先进的结果.

更多相关视频

Heterogeneity Mapping of Protein Expression in Tumors using Quantitative Immunofluorescence
07:54

Heterogeneity Mapping of Protein Expression in Tumors using Quantitative Immunofluorescence

Published on: October 25, 2011

19.1K
Quantitation of Protein Expression and Co-localization Using Multiplexed Immuno-histochemical Staining and Multispectral Imaging
08:40

Quantitation of Protein Expression and Co-localization Using Multiplexed Immuno-histochemical Staining and Multispectral Imaging

Published on: April 8, 2016

13.4K

相关实验视频

Last Updated: Jan 18, 2026

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
05:33

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System

Published on: July 11, 2025

809
Heterogeneity Mapping of Protein Expression in Tumors using Quantitative Immunofluorescence
07:54

Heterogeneity Mapping of Protein Expression in Tumors using Quantitative Immunofluorescence

Published on: October 25, 2011

19.1K
Quantitation of Protein Expression and Co-localization Using Multiplexed Immuno-histochemical Staining and Multispectral Imaging
08:40

Quantitation of Protein Expression and Co-localization Using Multiplexed Immuno-histochemical Staining and Multispectral Imaging

Published on: April 8, 2016

13.4K

科学领域:

  • 计算病理学计算病理学
  • 人工智能在瘤学中的应用
  • 数字病理学数字病理学

背景情况:

  • 从整个幻灯片图像 (WSIs) 准确的癌症诊断受到有限的注释,复杂的瘤结构和域移动的阻碍.
  • 现有的多实例学习 (MIL) 方法在计算病理学中与一般化和可解释性作斗争.

研究的目的:

  • 引入StructMIL,这是一个新的框架,用于从WSIs中进行可靠和可解释的癌症检测和分级.
  • 提高计算病理学模型的准确性,概括性和解释性.

主要方法:

  • 开发了一个结构意识,原型驱动的MIL框架 (StructMIL).
  • 基于集成图的拓先验和组织学上下文.
  • 采用原型增强的聚合来实现稳定的预测.
  • 实施了一个域泛化策略,包括对比对齐,对抗混和一致性规范化.

主要成果:

  • 在Camelyon16 (乳腺癌转移检测) 和PANDA (前列腺癌格里森分级) 上,StructMIL实现了最先进的性能.
  • 在Camelyon16 (AUC 0.967) 上,交叉中心AUC提高了3.2% (AUC 0.967).
  • 与之前的MIL模型相比,在PANDA上增加了7.4%的Gleason交叉扫描器评分强度.与之前的MIL模型相比,Cohen的Kappa在PANDA上增加了7.4%.
  • 生成可解释的基于原型的归因地图,突出显示有意义的结构.

结论:

  • 通过提高准确性,可解释性和概括性,StructMIL为多中心计算病理学工作流提供了一个实用的解决方案.
  • 该框架显示了对扫描仪和机构之间域名转移的稳定性的显著改进.
  • StructMIL为癌症诊断和分级提供可靠和可解释的见解.