Sørensen F B, Ottosen P D
Stereological Research Laboratory, University of Arrhus, Denmark.
Am J Dermatopathol. 1991 Apr;13(2):99-107. doi: 10.1097/00000372-199104000-00001.
The volume-weighted, mean nuclear volume (nuclear vv) may be estimated without any assumptions regarding nuclear shape using modern stereological techniques. As a part of an investigation concerning the prospects of nuclear vv for classification and malignancy grading of cutaneous melanocytic tumors, the observer variability of estimates of nuclear vv is studied. Routinely processed, paraffin embedded tissue specimens from 22 malignant melanomas and nine benign melanocytic cutaneous lesions are retrospectively investigated. The sampling scheme for estimation of nuclear vv is easy to use and robust, with more than 85% of the associated observed variance explained by differences among different tumors. Inter- and intraobserver reproducibilities are high, showing correlation coefficients of .86 and .96, respectively, with slopes of the regression lines close to unity. It is concluded that estimates of the three-dimensional nuclear vv in melanocytic cutaneous tumors are objective, unbiased, and highly reproducible.
使用现代体视学技术,无需对细胞核形状做任何假设,即可估计体积加权平均核体积(核vv)。作为关于核vv在皮肤黑素细胞肿瘤分类和恶性程度分级方面前景研究的一部分,对核vv估计值的观察者变异性进行了研究。回顾性调查了22例恶性黑色素瘤和9例良性皮肤黑素细胞病变的常规处理石蜡包埋组织标本。核vv估计的抽样方案易于使用且稳健,超过85%的相关观察方差可由不同肿瘤之间的差异解释。观察者间和观察者内的可重复性都很高,相关系数分别为0.86和0.96,回归线斜率接近1。得出结论,皮肤黑素细胞肿瘤中三维核vv的估计是客观、无偏且高度可重复的。