MR动态增强图像纹理分析判断乳腺结节良恶性的价值

目的 探讨MR动态增强图像纹理分析鉴别诊断乳腺结节良恶性的价值。方法 回顾性分析经手术病理证实的78例患者共80个乳腺结节的MR动态增强图像，每个结节获得63个纹理特征参数。绘制纹理参数鉴别诊断良恶性乳腺结节的ROC曲线，并与MR乳腺影像报告和数据系统（BI-RADS）的诊断效能比较。结果 78例患者的80个乳腺结节中，纹理参数中灰度游程长不均匀度判断乳腺结节良恶性的AUC值（0.836）最大且诊断准确率高，其诊断恶性乳腺结节的敏感度为82.93%（34/41）、特异度为94.87%（37/39）、准确率为88.75%（71/80）、阳性预测值为94.44%（34/36）、阴性预测值为84.09%（37/44）。MR BI-RADS分类诊断恶性乳腺结节的敏感度为95.12%（39/41）、特异度为87.18%（34/39）、准确率为91.25%（73/80）、阳性预测值为88.63%（39/44）、阴性预测值为94.44%（34/36）。MR BI-RADS分类和纹理分析判断恶性乳腺结节准确率差异无统计学意义（P=0.11）。与单独应用BI-RADS分类比较，两者联合应用可明显提高诊断恶性乳腺结节的特异度（P<0.001）。结论 MR纹理分析可作为传统诊断乳腺良恶性结节的补充。

Value of texture feature analysis of MRI dynamic contrast enhancement in diagnosis of benign and malignant breast nodules

Objective To assess the diagnostic value of texture analysis of MRI in differential dignosis of benign and malignant breast nodules. Methods The MRI data of 78 patients (80 breast nodules) identified by surgical pathology were retrospectively studied. Sixty-three texture parameters were obtained from each nodule. ROC curve of texture parameters in differential diagnosis of benign and malignant breast nodules were performed. Results In all of the 63 texture parameters, the run length nonuniformity (RLN) had the highest AUC value (0.836) and accuracy, the diagnostic sensitivity, specificity, accuracy, positive predictive value and negative predictive value in differentiation of breast nodules were 82.93% (34/41), 94.87% (37/39), 88.75% (71/80), 94.44% (34/36) and 84.09% (37/44). The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of breast imaging reporting and data system (BI-RADS) were 95.12% (39/41), 87.18% (34/39), 91.25% (73/80), 88.63% (39/44), and 94.44% (34/36). The difference of diagnostic accuracy between texture parameter and BI-RADS had no statistical significance (P=0.11). BI-RADS combined texture parameter improved specificity significantly (P<0.001). Conclusion The texture analysis could be complementary to improve the accuracy of BI-RADS-MRI in breast nodules.