Suppr超能文献

在非霍奇金淋巴瘤淋巴结切片形态计量描述中,两种用于细胞核形状和大小的统计方法。

Two statistical approaches to nuclei shape and size in a morphometric description of lymph node sections in non-Hodgkin's lymphoma.

作者信息

Lesty C, Raphaël M, Nonnenmacher L, Binet J L

机构信息

Département d'Hématologie CNRS UA 625, Centre de l'Association Claude Bernard, Centre Hospitalier Universitaire la Pitié-Salpêtrière, Paris, France.

出版信息

Cytometry. 1989 Jan;10(1):28-36. doi: 10.1002/cyto.990100106.

Abstract

An attempt is made to evaluate more clearly the potential contribution of quantitative nuclear profile shape and size measurements to lymph node section histologic description in 70 cases of non-Hodgkin's lymphoma (NHL). The area of nuclear profiles and five nonredundant and size-free shape indices were measured using the THECLA program on a Leitz Texture Analysis System (Leitz-TAS). Two statistical approaches were applied, known respectively as "parametric," (first statistical moments of variable distributions over samples of 200 nuclear profiles) and "nonparametric," which are percentages of nuclear profiles distributed into five "cytological" classes that are defined by shape: round (A), elongated (B), kidney shaped (C), irregular (D) and cleaved (E) nuclear profiles. Both statistical approaches provide proper overall discrimination of the eight histological categories identified with reasonable reliability by pathologists. Above all, the present report discusses the ability of a set of parametric and nonparametric variables to describe NHL cell populations, in an objective and meaningful way, according to nuclei shape. A method of synthesizing multidimensional correlations (CORICO program) is proposed in support of the discussion. Also, the specific descriptive power of each of the variables is described; in particular, it is concluded that there is a close link between the shape and size of the nuclear profiles of the cells.

摘要

我们试图更清晰地评估在70例非霍奇金淋巴瘤(NHL)中,定量核轮廓形状和大小测量对淋巴结切片组织学描述的潜在贡献。使用Leitz纹理分析系统(Leitz-TAS)上的THECLA程序测量核轮廓面积以及五个非冗余且无尺寸的形状指数。应用了两种统计方法,分别称为“参数法”(对200个核轮廓样本的变量分布的一阶统计矩)和“非参数法”,后者是根据形状分为五个“细胞学”类别的核轮廓的百分比,这五个类别分别为圆形(A)、细长形(B)、肾形(C)、不规则形(D)和裂开形(E)核轮廓。两种统计方法都能对病理学家以合理可靠性识别出的八个组织学类别进行恰当的总体区分。最重要的是,本报告讨论了一组参数和非参数变量根据细胞核形状以客观且有意义的方式描述NHL细胞群体的能力。为支持该讨论,提出了一种合成多维相关性的方法(CORICO程序)。此外,还描述了每个变量的具体描述能力;特别地,得出结论:细胞的核轮廓形状和大小之间存在密切联系。

相似文献

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