用于对子宫内腺癌的分类而进行的胰腺病理图像分析Silva模式依赖于弱监督的深度学习
Qingqing Liu1, Xiaofang Zhang2, Xuji Jiang1
1Cheeloo College of Medicine, Shandong University, Jinan City, China.
The American journal of pathology
|February 21, 2024
概括
一个新的AI工具Silva3-AI准确地从基因病理图像中分类了宫内腺癌的Silva模式. 这种深度学习管道显示了与经验丰富的病理学家可比的性能,有助于瘤分析.
科学领域:
- 在瘤学瘤学.
- 数字病理学数字病理学
- 人工智能的人工智能
背景情况:
- 宫内腺癌 (EAC) 占宫癌的25%,并且具有高度的异质性.
- 基于形态学的Silva模式系统用于风险分层,但难以准确分类.
- 瘤微环境的异质性影响Silva模式和临床结果.
研究的目的:
- 开发和验证深度学习管道Silva3-AI,用于对整个幻灯片图像进行自动分析.
- 为了准确地识别EAC中的Silva模式,使用基因病理图像.
- 与专家病理学家相比,评估Silva3-AI的性能.
主要方法:
- 使用视觉变压器和循环神经网络架构开发Silva3-AI模型.
- 培训和内部验证来自Qilu医院的202个EAC患者幻灯片.
- 在7个额外的医疗中心的161个EAC患者幻灯片上进行了独立测试.
主要成果:
- 在接收器运行的特征曲线值下,Silva3-AI实现了高类特定面积:在独立测试集上,Silva A为0.947,Silva B为0.908,Silva C为0.947.
- 人工智能模型的表现与具有10年诊断经验的病理学家的表现相当.
- 来自Silva3-AI的预测热图有助于可视化瘤微环境异质性.
结论:
- 席尔瓦3-AI在分类EAC席尔瓦模式方面表现出高精度,提供了可靠的自动化解决方案.
- 人工智能工具提供客观和可重现的分析,有可能改善风险分层和患者管理.
- 使用人工智能对组织病理学图像的自动分析可以增强对EACs瘤异质性的理解.
更多相关视频
相关概念视频
Classification of Epithelial Tissues: Overview
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,...
Based on the number of cell layers,...
Histology of the Uterus
The uterine wall consists of three histological layers: the perimetrium, myometrium, and endometrium. The outermost perimetrium is a thin, serous membrane connected with the broad ligament on the sides, which helps anchor the uterus in the pelvic cavity. The thickest layer, myometrium, is mainly made up of smooth muscle tissue bundles. Its contractions are vital in facilitating the expulsion of the uterine lining, fetus, and placenta during menstruation and childbirth.
The endometrium is the...
The endometrium is the...


