小肠平滑肌肿瘤病理分级的模糊模式识别

Fuzzy pattern discernment of small intestine leiomyoma in pathological classification

目的小肠平滑肌肿瘤的模糊数学研究．方法首先用模糊聚类分析（ＦＣＡ）把小肠平滑肌肿瘤３３５例分成４个子集．然后建立模糊模式识别（ＦＰＤ）方法．用常规病理诊断（ＲＰＤ）作对照．用ＦＣＡ的结果作标准．结果ＦＰＤ的符合率９５２％，而ＲＰＤ为５９７％．为了方便，编制了小肠平滑肌肿瘤的专用诊断软盘．结论ＦＣＡ的结果满意地克服了ＲＰＤ的一些弊病，但它不能处理日常工作中某一病例的诊断，所以必须采用ＦＰＤ．

IM To study fuzzy mathematics in small intestinal leiomyoma.METHODS 335 cases of small intestinal leiomyomas were classified into 4 sets by fuzzy cluster analysis (FCA). Then the method of fuzzy pattern discernment (FPD) was established. Routine pathologic diagnosis (RPD) served as controls. The results of FCA were used as standard.RESULTS The coincidence rate of FPD was 952%, and of RPD was 597%. A specialized diagnostic software for leiomyoma of small intestine was designed.CONCLUSION The result of FCA surmounted satisfactorily a few drawbacks of RPD, but ECA could not handle the diagnosis of a certain case in daily work, therefore, FPD must be adopted.