目的 优选出MRI诊断乳腺疾病的 诊断指标和扫描方案。方法 收集行MRI检 查的乳腺病例118例，并根据病理结果分 成良性病变和恶性病变两组，乳腺良性 病变68例，乳腺恶性病变50例；采用的 MRI序列包括T1加权成像、T2加权成像、 压脂的T2加权成像、扩散加权成像、动态 增强序列及T2*灌注成像；采用Logistic 回归分析优选出乳腺MRI检查的诊断指 标。结果 通过Logistic回归分析，肿 块的形态学特点、表观弥散系数值、时 间-信号强度曲线类型引入方程，并得 出回归方程为Logit(P)=0.280+1.919X2- 2.582X4+1.824X5。结论 肿物的形态特 点、表观弥散系数值、时间-信号强度 曲线类型有助于鉴别诊断乳腺良恶性疾 病；既满足诊断需要，同时最大限度缩短 扫描时间，建议精简乳腺MRI扫描方案为 T2W+T2W-SPAIR+DCE+DWI序列。

Objective To optimize MRI diagnosis indicator and examination scheme in breast disease. Methods Totally 118 cases with breast diseases were performed by MRI, and were classified into two groups according to the pathologic reports: 50 cases with breast cancers, and another 68 cases with benign breast tumor. MRI scanning sequence includes T1-weighted imaging, T2-weighted imaging, fat suppression of the T2-weighted imaging (T2W-SPAIR), diffusion weighted imaging (DWI), dynamic contrast enhanced (DCE) and T2* perfusion weighted imaging sequence. Logistic regression analysis was used to optimize the diagnosis indicators of breast MRI examination. Results Through Logistic regression analysis, the morphology feature of the mass, the apparent diffusion coefficient (ADC) value and the time-signal intensity curve (TIC) type were selected in the equation. The regression equation was Logit(P)=0.280+1.919X2-2.582X 4+1.824X5. Conclusion The morphology feature of the mass, the ADC value and the TIC type contribute to the differential diagnosis of benign and malignant breast disease. For both meeting the needs of the diagnosis in breast disease, and shortening the scan time of the breast MRI examination, T2W combine with T2W-SPAIR, DCE and DWI sequences are the simplification breast MRI examination scheme.