ADC纹理分析在鉴别浆细胞性乳腺炎和导管原位癌的价值

Value of image texture analysis of ADC in the differential diagnosis of plasma cell mastitis and ductal carcinoma in situ

目的探讨DWI ADC纹理分析在鉴别浆细胞性乳腺炎和导管原位癌的价值。方法选取我院46例经病理证实为浆细胞性乳腺炎和导管原位癌患者的常规MR图像和ADC图像,利用后处理软件对乳腺病变勾画感兴趣区(ROI),分别对其ADC图像进行纹理特征分析,并结合受试者工作特征曲线(ROC曲线)评估2组间纹理参数的鉴别诊断价值。结果统计得到纹理分析直方图中的平均值、方差、百分位数参数在2组数据中差异有统计学意义(P<0.05),应用ROC曲线及AUC得到第1百分位数及第10百分位数具有较高的鉴别诊断效能,AUC值分别为0.83、0.73。参考约登指数,第1百分位数取阈值为77.00时,敏感度与特异度分别为71.43%、81.82%;第10百分位数取阈值为161.5时,敏感度与特异度分别为80.95%、68.18%。结论浆细胞性乳腺炎与导管原位癌在ADC图像的纹理特征显著不同,ADC的纹理分析在两者鉴别方面具有可行性,可以作为鉴别两者的辅助方法。

Objective To investigate the value of apparent diffusion coefficient(ADC)texture parameters in evaluation of plasma cell mastitis and ductal carcinoma in situ.Methods The retrospective study was based on breast ADC images.A total of 46 patients with histopathologically confirmed plasma cell mastitis and ductal carcinoma in situ from Cangzhou People’s Hospital were enrolled in this study.The textural features of ADC images were analyzed.The receiver operating characteristic curve(ROC curve)was calculated to evaluate the diagnostic efficiency of texture parameters.Results The statistical results showed that the mean value,variance and percentile parameters in the histogram of texture analysis had statistical significance in the two groups of data(P<0.05).The application of ROC curve and AUC to get the Pere.1%and Pere.10%had a high differential diagnosis efficiency,and the AUC value was 0.83 and 0.73,respectively.According to Youden index,when the threshold value of the Pere.1%was 77.00,the sensitivity and specificity was 71.43%and 81.82%respectively.When the Pere.10%threshold was 161.5,the sensitivity and specificity were 80.95%and 68.18%,respectively.Conclusion The texture characteristics on ADC images of plasma cell mastitis and ductal carcinoma in situ are significantly different.Texture analysis based on ADC images is feasible in identifying them,It can be used as an auxiliary method to identify them.