乳腺癌患者磁共振扩散加权成像感兴趣区的选择与生物预后因子的相关性分析

Correlation between ROI Selection and Biological Prognostic Factors in Diffusion Weighted Imaging of Breast Cancer

目的评价磁共振扩散加权成像(DWI)中感兴趣区(ROI)的三种不同选择与乳腺癌生物预后影响因子的相关性及其应用价值。方法回顾性分析经手术并病理诊断为乳腺癌的82例患者资料。根据术后病理证实将乳腺癌分为雌激素受体(ER)阳性组及阴性组、孕激素受体(PR)阳性组及阴性组、人表皮生长因子受体-2(HER-2)阳性组及阴性组。两名高低年资放射医师独立在乳腺DWI图像中选取三种不同ROI法(小ROI、面积ROI、容积ROI)测量病灶的表观扩散系数(ADC)值。结果三种选取ROI方法所测得的ADC值可重复性均较高,组内相关系数(ICC)>0.90。三种测量方法的ADC值在PR及ER阳性及阴性组之间有差异,ER、PR阳性组的ADC低于阴性组。三种测量方法测得的ADC值在HER-2阳性组及阴性组之间无明显差异。ER受体表达中容积ROI法最佳临界值为0.962×10^(-3)mm^(2)/s,曲线下面积(AUC)为0.724;PR受体表达中容积ROI最佳临界值为0.929×10^(-3)mm^(2)/s, AUC为0.803。三种测量方法的ADC值与HER-2受体的表达相关性不明显。结论三种ROI法测得ADC值能用于预测PR与ER情况,其中容积ROI法测得的ADC值在预测PR方面诊断效能最高。

Objective To evaluate the correlation between three different methods of diffusion weighted imaging(DWI) and early prognostic factors of breast cancer. Methods 82 cases of breast cancer were analyzed retrospectively.According to the postoperative pathology, breast cancer was divided into ER positivity and negative, PR positivity and negative and HER-2 positivity and negative.Three different ROI methods(small ROI,area ROI and volume ROI) were used to measure the ADC value of breast lesions in DWI images by the two radiologists. Results The repeatability of ADC values measured by the three selected ROI methods was high.The ADC values of the three methods were different between PR and ER positive and negative groups.The ADC values of the high expression group of ER and PR were lower than that of the low expression group.There was no significant difference in ADC values between HER-2 positive and HER-2 negative groups.The best critical value was 0.962×10^(-3)mm^(2)/s and the area under the receiver operating characteristic curve(AUC) was 0.724;The best critical value was 0.929×10^(-3)mm^(2)/s and the AUC was 0.803.There was no significant correlation between ADC values and HER-2 receptor expression. Conclusion The ADC values measured by the three ROI methods could be used to predict PR and ER,and the ADC values measured by the volume ROI method had the highest diagnostic efficiency in predicting PR.