中文摘要：

目的 评价3.0T MR多个不同b值扩散加权成像中的表观弥散系数（ADC）对前列腺癌的诊断价值。方法 回顾性分析52例经病理证实为前列腺疾病的患者,包括前列腺癌（PCa组）28例和前列腺增生（BPH组）24例,并以20例正常前列腺MRI作为对照（正常组）。所有患者及对照者均行核磁共振（MRI）常规扫描序列（TIWI、T2WI）和多b值DWI检查（b值=0、500、1 000、1 500、2 000、2 500s/mm2）。统计并比较不同b值时前列腺良、恶性组织及正常前列腺之间ADC值的差异,通过受试者工作特征（ROC）曲线分析,确定具有最大ROC曲线下面积（AUC）的b值,并计算相应诊断界值、敏感性和特异性。结果 不同b值DWI PCa组与BPH组、PCa组与正常组平均ADC值之间,差异均有统计学意义（P〈0.05）,但BPH组与正常组平均ADC值之间,差异无统计学意义（P〉0.05）。当b值为2 000s/mm2时,AUC最大（0.963）,其对应的诊断界值、敏感性、特异性分别为0.847×10-3 s/mm2、100.0%、86.2%。结论 高b值（b=2 000s/mm2）DWI图像对前列腺癌的诊断准确性最高,其对应的诊断界值有助于前列腺疾病的鉴别诊断,并具有较高的敏感性和特异性。

英文摘要：

Objective To evaluate the clinical diagnostic value of prostate cancer under multiple-b-value diffusion weighted MRI, and determine the optimal b value,diagnostic cut-off value, sensitivity and specificity. Methods 52 patients with prostate disease were retrospectively analyzed in the study,including 28 cases of prostate cancer（PCa） and 24 cases of benign prostate hyperplasia （BPH）. In addition,20 volunteers of non-prostate disease. All cases and volunteei＇s were performed traditional MRI（T1WI,T2WI） and DWI（b= 0,500,1 000,1 500,2 000,2 500 s/mm2 ） examination. ADC values were calculated and compared among prostate benign tissues,malignant tissues and normal tissue of multiply different b values, the biggest AUC（area under the curve）was determined by ROC curve, the opposite sensitivity and specificity were calculated. Results Under every different b-value diffusion weighted MRI,the differences of the mean ADC values between PCa and BPH, PCa and normal prostate were statistically significant（P〈0.05）. But difference between BPH and normal prostate had no statistically Significant（P〉0. 05）. Under b= 2 000 s/mm2,the AUC was biggest （0. 963）. The corresponding diagnostic cut-off value, sensitivity and specificity was 0. 847 × 10^-3 s/mm^2 , 100.0%,86.2 %. Conclusion DWI（b= 2 000 s/mm^2） has higher diagnostic accuracy of PCa. It＇s corresponding diagnostic cut-off value is helpful to differentiate diagnose prostate disease,and has a high degree of sensitivity and specificity.