摘要
目的：探索T2 mapping成像在乳腺病变良恶性鉴别诊断中的应用价值。方法：收集乳腺癌病灶13个,乳腺良性病灶13个,行乳腺磁共振T2 mapping检查,分别测量每个病灶T2值的均值（μ）及标准差（σ）,并计算每个病灶T2值的离散系数（CV,CV=σ/μ）。对两组患者病灶μ及CV分别进行Student’s t检验。应用受试者工作特征（ROC）曲线评价乳腺病变T2值的μ及CV对乳腺癌的诊断价值,确定诊断截点。结果：乳腺癌与乳腺良性病变的μ值间无显著统计学差异（t=1.164,P=0.256）;乳腺癌与乳腺良性病变T2值的CV间具有显著的统计学差异（t=6.203,P=0.000）。以乳腺病变T2值的CV作为诊断指标进行ROC分析,曲线下面积（AUC）为0.964（AUC〉0.9,有较高的诊断价值）,截点为0.145,以此值判断乳腺癌,敏感度92.3%（12/13）,特异度100%（13/13）,准确度0.962（25/26）。以乳腺病变T2值的μ作为诊断指标进行ROC分析,AUC为0.627（AUC介于0.5~0.7,有较低的诊断价值）。结论：在T2 mapping成像中,乳腺恶性病灶的T2值的CV高于良性病灶。T2 mapping成像在乳腺病变良恶性鉴别诊断中具有一定的应用价值。
Objective： To evaluate the usefulness of MR T2 mapping in differentiating breast carcinoma from benign lesions.Materials and Methods： MR T2 mapping data was obtained from 13 lesions of breast carcinoma and 13 lesions of breast benign lesions.The mean value（μ） and standard deviation（σ） of T2 values in the lesions were measured.The coefficient of variance（CV） of the lesion was defined as CV=σ/μ.The μ values and CVs were compared in the two groups.The receiver operating characteristic（ROC） curve was used to evaluate the diagnostic value of μ values and CVs of T2 values for breast carcinoma,and then determine the cut-off point.Result： There were no obvious differences between the mean value of breast carcinoma and breast benign lesions（t=1.164,P=0.256）.There were statistically difference between CVs of breast carcinoma and breast benign lesions（t=6.203,P=0.000）.When ROC analysis of the CVs of T2 values of breast carcinoma as a diagnostic indicator,area under the curve（AUC） is 0.964（AUC0.9,high diagnostic value）,as the cut-off point was 0.145,the diagnosis of breast cancer with this standard,sensitivity is 92.3%（12/13）,specificity is 100%（13/13）,and the accuracy is 0.962（25/26）.When ROC analysis of μ values of T2 of breast carcinoma as a diagnostic indicator,AUC is 0.627（AUC is between 0.5 and 0.7,low diagnostic value）.Conclusion： In MR T2 mapping imaging,CV of T2 values of breast malignant lesions is higher than that of benign lesions.The CV of T2 values of breast lesions is more useful than the mean value in distinguishing breast carcinoma from breast benign lesions.
出处
《中国临床医学影像杂志》
CAS
2013年第5期317-320,共4页
Journal of China Clinic Medical Imaging
关键词
乳腺肿瘤
乳腺疾病
磁共振成像
Breast neoplasms
Breast diseases
Magnetic resonance imaging
