同时多层分段读出平面回波弥散加权成像鉴别乳腺良恶性肿瘤  

Differentiation of benign and malignant breast tumors with simultaneous multislice readout segmented echo planar diffusion weighted imaging

作　　者：李雪 严福华[1] 柴维敏[1] 朱宏[1] 傅彩霞 刘薇 孙琨[1] LI Xue;YAN Fuhua;CHAI Weimin;ZHU Hong;FU Caixia;LIU Wei;SUN Kun(Department of Radiology,Ruijin Hospital,Shanghai Jiao Tong University School of Medicine,Shanghai 200025,China;Application Development,Siemens Shenzhen Magnetic Resonance Ltd.,Shenzhen 518057,China)

机构地区：[1]上海交通大学医学院附属瑞金医院放射科,上海200025 [2]西门子(深圳)磁共振有限公司应用开发部,广东深圳518057

出　　处：《中国医学影像技术》2021年第8期1224-1228,共5页Chinese Journal of Medical Imaging Technology

摘　　要：目的评价同时多层(SMS)分段读出平面回波(RS-EP)弥散加权成像(DWI)用于乳腺的可行性及全肿瘤直方纹理分析参数鉴别乳腺良、恶性肿瘤的价值。方法纳入185例经病理证实的女性乳腺肿瘤患者,均为单发病灶,根据肿瘤性质分为良性组(n=58)及恶性组(n=127);采集SMS RS-EP DWI图像,提取全肿瘤直方图参数及纹理参数,采用组内相关系数(ICC)评价观察者内及观察者间的可重复性,并对比组间各参数差异。绘制受试者工作特征(ROC)曲线,计算曲线下面积(AUC)评价定量参数鉴别乳腺良、恶性肿瘤的效能。结果观察者内及观察者间ADC Kurtosis、观察者间ADC Entropy一致性中等(ICC均≤0.75),其余参数均一致性较好或极好(ICC均>0.75)。除ADC SD和ADC Kurtosis外,其他参数组间差异均有统计学意义(P均<0.05)。直方图参数中,ADC Median鉴别乳腺良、恶性肿瘤的AUC最高(0.879);纹理参数中,ADC Entropy的AUC值最高(0.764);二者联合诊断的AUC为0.911,敏感度和特异度分别为90.55%和81.03%。结论SMS RS-EP DWI可用于乳腺检查;联合全肿瘤直方纹理分析有助于鉴别乳腺良、恶性肿瘤。Objective To observe the feasibility of simultaneous multislice(SMS)readout segmented echo planar(RS-EP)diffusion weighted imaging(DWI)in examinations of the breast,and to observe the value of whole-tumor histogram and texture analysis parameters for differentiating benign and malignant breast tumors.Methods A total of 185 female patients with single tumor of breast confirmed by pathology were enrolled and were divided into benign group(n=58)and malignant group(n=127)according to the nature of the lesions.The prototype SMS RS-EP DWI sequence scanning were performed,and the whole-tumor-based histogram and texture parameters were extracted from ADC maps based on a whole-tumor analysis.Intra-class correlation coefficient(ICC)was used to evaluate the intra-observer and inter-observer repeatability,and the parameters were compared between groups.Receiver operating characteristic(ROC)curves were drawn and the area under the curve(AUC)was calculated to evaluate the capability of the quantitative parameters to discriminate breast lesions.Results The intra-and inter-class observer agreement of ADC Kurtosis and the inter-class observer agreement of entropy were of moderate reliability(all ICC≤0.75),good-to-excellent agreements were obtained between the observers for the other parameters(all ICC>0.75).Except for ADC SD and ADC Kurtosis,the other parameters were all statistically different between groups(all P<0.05).ADC Median had the highest AUC(0.879)among histogram parameters,while ADC Entropy had the highest AUC(0.764)among texture parameters.The AUC of combining of ADC Median and ADC Entropy for differentiating benign and malignant breast tumors was 0.911,and the sensitivity and specificity was 90.55%and 81.03%,respectively.Conclusion SMS RS-EP DWI could be used for breast scanning,and combining with straight texture analysis of the whole tumor were helpful to differentiating benign and malignant breast tumors.

关 键 词：乳腺肿瘤 弥散磁共振成像 

分 类 号：R737.9[医药卫生—肿瘤] R445.2[医药卫生—临床医学]