多元判别分析在颌骨骨肉瘤病理诊断中的应用

本文用交互式图象分析系统测定了46例颌骨骨肉瘤的细胞核形态和DNA含量,并对测定参数进行了多元判别分析,以探讨其对颌骨骨肉瘤组织学类型和分级的诊断价值。结果表明对组织类型鉴别能力较大的参数依次细胞核最大径标准差和DNA含量平均值等;判断正确率为71.74％;对组织学分级判别作用较大的参数依次为细胞核最小径标准差和细胞核最大径平均值等,判断正确率为89.31。本实验结果可为临床病理诊断骨肉瘤组织类型和分级提供参考指标。

Application of Multiple Discriminative Analysis in Pathologic Diagnosis of Osteosarcoma of the Jaws

The measurements of the cell nuclear morphology and the quantitative analysis of the nuclear DNA content were made using Interactive Image Analysis System in 46 patients with osteosarcoma of the jaws. The measured parameters were studied with multiple discriminative analysis. The present paper sought to deterrain whether this method may be of the value to diagnosing the histologic type and grade of osteosarcoma. The results showed that standard deviation of thlargest diameter and the smallest diameter of the cell nuclei were the most effeective parameters to discriminat the histologic type and grade, respectively. The corrective rates of discrimination were 71.74 per cent for histologie type, and 89.13 per cent for grade This study may be helpful to provide referenec indexes of diagosing the histologic type and grade of osteosarcoma