人乳腺病变组织AgNORs的形态定量分析

目的：探讨计算机图像分析技术对乳腺肿瘤核仁组成区相关嗜银蛋白（AgNORs）表达在病理学诊断中的意义。方法：选取乳癌、乳腺结构不良、乳腺纤维腺瘤及正常乳腺组织各20例，应用图像分析系统和嗜银染色技术，对乳腺病变组织的AgNORs进行形态计量学分析，观察每个细胞的胞核面积、核周长以及AgNOR颗粒数目、颗粒面积、AgNOR的大颗粒数、形态因子、面积比值、平均面积等8项参数。结果：正常乳腺及乳腺结构不良、乳腺纤维腺瘤组织的细胞核面积、核周长以及AgNOR颗粒数目、颗粒面积明显小于乳癌组织（P＜0．01），AgNOR的大颗粒数、形态因子、面积比值、平均面积在乳癌组织与乳腺纤维腺瘤及正常乳腺组织间差异有统计学意义（P＜0．05）。上述8项参数4组间多类判别分析回代符合率为72．5％，癌和非癌的判别回代符合率达100％。结论：图像分析对于乳腺良、恶性病变的鉴别诊断具有重要价值，是一精确、客观的AgNOR定量研究手段。

Morphometrical study on AgNORs of human breast lesions

Aim: To study the significance of image analysis in the argyrophilic protein associated with the nucleolar organizer regions (AgNORs) in pathological diagnosis of breast tumors. Methods: An image analysis system and silver staining techniques were applied to measure the nucleus and AgNORs in nuclei of human breast lesions. The morphometric parameters were mean area and perimeter of cell nucleus, averages of area and number of AgNORs per nucleus, mean area per AgNOR, the number of bigger AgNORs per nucleus, the ratio of AgNORs area to nucleus area per cell and mean form factor of AgNORs. Results:The morphometric parameters such as mean area and perimeter of cell nucleus, averages of area and number of AgNORs per nucleus, in normal, dysplastic,and fibroadenoma cells were significantly lower than that in breast cancer cells ( P <0.01). The mean area per AgNOR, the number of bigger AgNORs per nucleus and the ratio of AgNORs area to nucleus area per breast cell were lower in normal,dysplasia, and fibroadenoma than that in cancer cell ( P < 0.05). The mean form factor of AgNORs was decreased in cancer cell ( P <0.05). All predicated cells were correctly identified (reclassification) in cancer and non cancer cells. Conclusion: The image analysis is of important value in differentiating benigh breast lesions and malignant breast lesions,and is an accurate and reliable way of quantitative study on AgNORs.