Baak J P
Anal Quant Cytol Histol. 1987 May;9(2):89-95.
This article gives an overview of the quantitative pathologic techniques used today, with special emphasis on interactive morphometry and its application in the clinical setting. At present, stereologic calculations may be necessary, but in a diagnostic setting have only rarely proved to be essential. The reproducibility of the measurements is discussed in relation to the definition of the particles and staining methods. A number of technical factors that cause random errors are mentioned, such as quality of the slides, magnification, definition of the particles to be measured and measuring protocol. Large-scale experiments have revealed that the means of nuclear morphometric and certain stereologic features are reproducible, such as in the volume percentage of epithelium, the surface densities of glands and the mitotic activity index. The diagnostic applications of the standard deviation and shape factors of nuclear quantitative features require additional precautions, however. Having quantified cell and tissue features, multivariate analysis may result in a better discrimination of two or more groups under study. The quantitative pathologic examination of cells and tissues can provide important diagnostic and prognostic information. Quantitation in pathology is especially useful in so-called continuous lesions, in which interobserver and intraobserver disagreement is considerable. An important requirement of diagnostic morphometry is object selection by a skilled pathologist; the use of morphometry as a black box can result in dramatic errors. The criteria used for a morphometric classification rule that can be used for clinical applications are summarized. Quality control of the whole measuring system is essential. Application of these techniques for more than six years in diagnostic pathology has repeatedly corrected previous qualitative diagnoses.(ABSTRACT TRUNCATED AT 250 WORDS)
本文概述了当今使用的定量病理技术,特别强调了交互式形态计量学及其在临床环境中的应用。目前,立体学计算可能是必要的,但在诊断环境中很少被证明是必不可少的。测量的可重复性与颗粒的定义和染色方法相关进行了讨论。提到了一些导致随机误差的技术因素,如载玻片质量、放大倍数、待测量颗粒的定义和测量方案。大规模实验表明,核形态计量学的均值和某些立体学特征是可重复的,如上皮细胞的体积百分比、腺体的表面密度和有丝分裂活性指数。然而,核定量特征的标准差和形状因子的诊断应用需要额外的注意事项。在对细胞和组织特征进行量化后,多变量分析可能会更好地区分两个或更多正在研究的组。细胞和组织的定量病理检查可以提供重要的诊断和预后信息。病理学中的定量在所谓的连续性病变中特别有用,在这些病变中,观察者间和观察者内的分歧相当大。诊断形态计量学的一个重要要求是由熟练的病理学家进行对象选择;将形态计量学用作黑箱可能会导致严重错误。总结了可用于临床应用的形态计量分类规则所使用的标准。整个测量系统的质量控制至关重要。这些技术在诊断病理学中应用六年多来,多次纠正了以前的定性诊断。(摘要截短至250字)