中文摘要：

癌巢的侵袭行为是乳腺浸润性导管癌（invasive ductal carcinoma,IDC）恶性生物学行为过程的主要体现。通过免疫组织化学方法特异性标记乳腺癌癌巢,并利用图像处理与分析技术,自动从肿瘤图像中提取IDC癌巢的数理参数,研究其与乳腺癌患者无病生存时间（disease freesurvival,DFS）的关系。提取8个数理参数,经多因素分析显示,癌巢总周长（P=0.043）和圆形度（P=0.015）是患者DFS的独立影响因子,其风险比分别为1.518[95%CI,1.013-2.273]和1.682[95%CI,1.107-2.555],高于淋巴结分期风险比（1.361[95%CI,1.085-1.706]）,低于组织学分级风险比（4.685[95%CI,2.865-7.660]）。结果表明,分析癌巢数理参数,有助于预测无淋巴结转移乳腺癌患者的预后信息。

英文摘要：

The invasive behaviors of cancer nests are the main reflects of invasive ductal carcinoma（IDC）’s malignant biological behavior.In present work,the authors adopted the image processing method to extract mathematical parameters of IDC cancer nests which were specifically stained by immunohis to chemical method.The relationship between cancer nests’ parameters and patients’ disease free survival was studied.There were 8 mathematical parameters extracted.Multivariate analysis showed that total perimeter（P=0.043） and circularity（P=0.015） of cancer nests were independent impact factors of patients’ disease free survival.The hazard risk ratios of total perimeter and circularity were 1.518（95% CI [1.013-2.273]） and 1.682（95% CI [1.107-2.555]） respectively,which were higher than that of lymph node staging（1.361,95% CI [1.085-1.706]） and lower than that of histological grading（4.685,95% CI [2.865-7.660]）.Results of this study indicated that analysis on mathematical parameters of cancer nest is helpful to predict the prognostic information of early stage breast cancer patients without node metastasis.