表观扩散系数值离散系数在乳腺良恶性病变鉴别诊断中的应用

摘要

目的:探讨表观扩散系数(ADC)值的离散系数对于乳腺良恶性病变鉴别诊断的应用价值.方法:搜集96例患者磁共振乳腺病灶98个,分别测量每一病灶ADC值的均值(x)及标准差(s),并计算每一个病灶ADC值的离散系数CV(CV=s/x).评估病灶ADC值均值及CV的诊断效能.结果:乳腺良恶性病变组ADC值均值[良性(1.42±0.32)×10-3mm2/s,恶性(1.04±0.29)×10-3mm2/s]及CV(良性0.10±0.03,恶性0.17±0.04),两组间均有统计学差异(P＜0.0001).乳腺良性与恶性病变之间ADC-CV差异的效应量(Cohen's d值,1.98)高于良性与恶性病变之间ADC均值差异的效应量(1.25).ADC-CV ROC曲线的AUC (0.935±0.0227)也显著大于ADC均值ROC曲线的AUC (0.790±0.0447)(P＜0.01).结论:乳腺病变DWI检查中ADC-CV较ADC值均值在诊断乳腺良恶性病变中具有更高的诊断效能.%Objective:To evaluate the usefulness of coefficient variance (CV)of apparent diffusion coefficient (ADC) value in differential diagnosis of benign and malignant breast lesions.Methods:The mean value (x) and standard deviation (s) of ADC value of 98 lesions in 96 patients were measured.The CV of the lesion was denned as CV=s/x.The efficiencies of ADC value and CV were assessed.Results:There was obvious difference between the mean value of ADC [benign (1.42 ±0.32) × 10-3 mm2 /s,malignant (1.04±0.29)× 10-3mm2/s] and CV (benign 0.10±0.03,malignant 0.17±0.04) of breast lesions (P＜0.0001).The Cohen's d value of the differentiation of ADC-CV between benign and malignant lesions (1.98) was higher than that of ADC value (1.25).And the AUC of ROC of CV (0.935±0.0227) was significantly greater than that of ADC value (0.790±0.0447)(P＜0.01).Conclusion:Compared with ADC value,ADC-CV can improve the diagnostic performance of breast lesions.