基于体素内不相干运动DWI在乳腺良恶性病变鉴别中的价值  

The Value of DWI Based on Intra Voxel Incoherent Motion in Identifying Benign and Malignant Breast Lesions

作　　者：王高燕[1] 杨光钊[2] 董海波[2] 徐凌斌 王波[2] 李晖[2] 

机构地区：[1]浙江大学医学院附属第二医院,315040 [2]宁波市医疗中心李惠利医院影像科

出　　处：《临床放射学杂志》2016年第3期348-352,共5页Journal of Clinical Radiology

摘　　要：目的探讨双指数模型DWI对乳腺良恶性病变的鉴别诊断价值。方法 45例患者行多b值DWI检查(12个b值:0、10、20、30、50、100、200、400、800、1200、1500、3000 s/mm^2)。22例良性(23个病灶),23例恶性。获得双指数模型参数:标准表观扩散系数ADC_(st)、慢速表观扩散系数ADCslow、快速表观扩散系数ADCfast、快速扩散成分所占比例ffast。绘制ROC曲线,比较各参数对乳腺良恶性病变的诊断效能。结果恶性组ADC_(st)(0.747±0.145)×10^(-3)mm^2/s、ADCslow(0.585±0.132)×10^(-3)mm^2/s、ffast(0.307±0.112),明显低于良性组;ADCfast(4.654±1.097)×10^(-3)mm^2/s高于良性组,差异无统计学意义。ADC_(st)、ADCslow、ADCfast、ffast的ROC曲线下面积(AUC)分别为0.905、0.917、0.796、0.756。判断乳腺良恶性病变的敏感度分别为73.9%、91.3%、82.6%、56.5%,特异度为95.6%、78.3%、60.9%、95.6%。结论乳腺肿瘤ADC值的下降是由于快速扩散成分所占比例的减少和慢速扩散成分所占比例的增加所致,而不是ADCfast值自身的减少所致。ADCslow在良恶性病变鉴别中最有价值。Objective To explore the IVIM Biexponential model of diffusion weighted imaging （DWI） and its value in identification of benign and malignant breast lesions. Methods 45 cases of patients were examined with multiple b-values of DWI （ 12 b values ： 0, 10,20,30,50,100,200,400,800,1200,1500,3000 s/mm2 ）. The cases comprised of benign ca- ses in 22 patients （23 lesions ） and malignant ones in 23 cases. We obtained the IVIM Biexponential model parameters （ standard apparent diffusion coefficient, ADCst ; slow apparent diffusion coefficient, ADCslow ; fast apparent diffusion coefficient, ADCfast; fraction of ADCfast, ffast）- Using the ROC curve we compared various parameters in the diagnosis of benign and malignant breast lesions. Results The malignant group＇ s ADCst（0.747±0. 145）×10^-3mm^2/s, ADCslow （0. 585±0. 132）×10^-3mm2/s, ffast（0. 307±0. 112） were significantly lower than the benign group. On the other hand, the malig- nant group＇ s ADCfast （4. 654±1. 097 ）×10^-3 mm2/s, was higher than the benign group but there was no statistically signif- icant difference. The calculated areas （AUC） of ADCst, ADCslow, ADCfast, fast, under the ROC curve were 0. 905,0. 917, 0. 796,0. 756 respectively. The sensitivity values to determine breast benign and malignant lesions were 73.9% ,91.3% , 82.6% ,56.5% , and the specificity values of breast benign and malignant lesions were95.6% ,78.3% ,60.9% ,95.6%. Conclusion The main reason for the decreasing ADC of a breast tumor is the decreasing fraction of the fast component and the increasing fraction of the slow component, rather than the decreasing ADC fast of the fast component itself. ADCslow is the most valuable parameter in the identification of benign and malignant lesions.

关 键 词：扩散加权成像 表观扩散系数 双指数模型 乳腺肿瘤 

分 类 号：R445.2[医药卫生—影像医学与核医学] R737.9[医药卫生—诊断学]