DWI联合DCEMRI对乳腺非肿块样强化病灶的诊断价值

DWI combined with DCE MRI for differential diagnosis between benign and malignant diseases in breast lesions with MRI none-mass enhancement

目的探讨弥散加权成像(diffusion-weighted MR imaging,DWI)联合动态对比增强磁共振成像(dynamic contrast-enhanced M R imaging,DCE M RI)对表现为非肿块样强化(none-mass enhancement,NM LE)的乳腺良恶性病变的诊断价值。方法回顾性分析49个乳腺NMLE病灶,比较良恶性组间强化分布、内部强化特征、动态曲线模式及表观弥散系数(apparent diffusion coefficient,ADC)值的差异有无统计学意义。绘制接受者操作特征曲线(receiver operating characteristic curve,ROC)评估ADC值的诊断效能、确定最佳诊断分界点。应用多因素分析探讨表现为NMLE的乳腺恶性病变的磁共振影像特征。结果良恶性组间病灶的内部强化特征、动态曲线模式的差异无统计学意义(P>0.05)。段样分布(P=0.01)、病灶内ADC值<1.2×10-3mm2/s(P<0.001)提示其为乳腺癌。段样分布的诊断灵敏度(sensitivity,Se)、特异度(specificity,Sp)、阳性预测值(positive predictive value,PPV)、阴性预测值(negative predictive value,NPV)分别为73.1%、87.0%、86.4%、74.1%。以ADC阈值为诊断标准,其Se、Sp、PPV、NPV分别为:84.6%、91.3%、91.7%、84.0%。当NMLE表现为段样分布且病灶内ADC值小于1.2×10-3mm2/s时强烈提示其为乳腺癌,PPV为92.9%。结论 DWI联合DCE M RI对于表现为NM LE的乳腺癌具有较高的诊断准确性。

Objective To evaluation the application of a combination of diffusion-weighted imaging（ DWI） and dynamic contrast-enhanced MR imaging（ DCEMR I） in differentiating benign from malignant lesions showing non-mass-liked enhancement（ NMLE） on breast MR imaging.Methods MRI findings of 49 breast lesions showing NMLE were retrospectively reviewed. The distribution,internal enhancement,kinetic curve pattern and minimal apparent diffusion coefficient（ ADC） values were analyzed. The threshold of ADC for differentiating lesions was determined with receiver operating characteristic（ ROC） analysis. The indicators for malignancy were determined with multivariate analysis. Results There was no significant difference in internal enhancement and kinetic curve pattern between benign and malignant lesions. Segmental distribution（ P〈 0. 001） and ADC value 1. 2 × 10^- 3mm2/ s were the MR indicators of malignancy. When lesions showsegmental distribution,the sensitivity,specificity,negative predictive value（ NPV） and positive predictive value（ PPV） were 73. 1%,87. 0%,86. 4% and 74. 1%,respectively. The ADC threshold yielded sensitivity（ 84. 6%）, specificity（ 91. 3%）, NPV（ 91. 7%） and PPV（ 84. 0%）. The segmental distribution of NMLE lesion with ADC value 1. 2 × 10^- 3mm2/ s strongly indicated the malignancy,with a PPV of 92. 9%. Conclusion The combination of diffusion-weighted imaging and dynamic contrast-enhanced MR imaging shows high diagnostic accuracy for breast NMLE lesions.