Suppr超能文献

子宫内膜刷检样本中正常、增生和恶性子宫内膜细胞的数字成像分析

Digital imaging analysis of normal, hyperplastic and malignant endometrial cells in endometrial brushing samples.

作者信息

Fu Y S, Ferenczy A, Huang I, Gelfand M M

机构信息

Department of Pathology, Jonsson Comprehensive Cancer Center, University of California, Los Angeles 90024-1732.

出版信息

Anal Quant Cytol Histol. 1988 Apr;10(2):139-49.

PMID:3390290
Abstract

Sixty cytologic specimens obtained by endometrial brushing (using the Gynecyte device) were quantitated by digital imaging techniques. These samples included 25 from normal endometria, 6 from persistent proliferative endometria, 14 from cystic or adenomatous hyperplasias and 15 from carcinomas. The morphometric parameters surveyed included mean cell area, nuclear area, perimeter and long and short axes. The amount of hematoxylin dye in the nuclei was expressed by mean transmittance (mean of gray levels) and chromatin index (standard deviation of gray levels). The frequency distributions of cells derived from normal tissue and persistent proliferative endometrium were quite similar. The quantitative parameters of cystic and adenomatous hyperplasia, although intermediate between those of normal endometrium and carcinoma, were closer to the former than to the latter. Using stepwise discriminant analysis of the morphometric parameters, 83% of the specimens were correctly classified into the categories of normal/persistent proliferative, hyperplasia and carcinoma. The accuracy was improved to 88% when densitometric parameters were added. This study demonstrates the potential application of digital imaging techniques to the distinction and classification of normal, hyperplastic and malignant endometrial cells.

摘要

通过子宫内膜刷检(使用Gynecyte装置)获取的60份细胞学标本,采用数字成像技术进行定量分析。这些样本包括25份来自正常子宫内膜的标本、6份来自持续性增殖性子宫内膜的标本、14份来自囊性或腺瘤样增生的标本以及15份来自癌组织的标本。所测量的形态学参数包括平均细胞面积、核面积、周长以及长轴和短轴。细胞核中苏木精染料的含量通过平均透光率(灰度级平均值)和染色质指数(灰度级标准差)来表示。来自正常组织和持续性增殖性子宫内膜的细胞频率分布非常相似。囊性和腺瘤样增生的定量参数虽然介于正常子宫内膜和癌组织之间,但更接近前者。通过对形态学参数进行逐步判别分析,83%的标本被正确分类为正常/持续性增殖、增生和癌类别。当加入光密度参数时,准确率提高到了88%。本研究证明了数字成像技术在正常、增生和恶性子宫内膜细胞的鉴别和分类中的潜在应用。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验