摘要
目的 评价ADC值及相对ADC值（rADC）在乳腺结节病灶检查中的应用价值.方法 对52例66个乳腺结节病灶行MR DWI扫描,b值分别为0、800、1000 s/mm^2,测量乳腺结节及同侧、对侧乳头层面未受累乳腺实质组织ADC值,计算rADC1（病灶ADC值/同侧未受累乳腺实质ADC值）及rADC2（病灶ADC值/对侧未受累乳腺实质ADC值）,对所获资料采用两独立样本t检验、x2检验进行统计学分析.所有病灶均经穿刺活检或手术病理证实.结果 52例患者中,浸润性导管癌18例,纤维腺瘤34例,DWI显示50例,共64个病灶.b=800s/mm^2时,良、恶性结节平均ADC值、同侧rADC800-1及对侧rADC800-2分别为（1.54±0.28）×10^-3、（1.01±0.09）×10^-3 mm^2/s和0.77±0.15、0.52±0.07,0.76±0.14、0.51±0.06；其差异均有统计学意义（t值分别为8.217、9.339、10.394,P值均＜0.01）；以乳腺浸润性导管癌平均值95％参考值范围上限界值作为恶性病变上限阈值点,ADC值、rADC8o0-1及rADC800-2分别为1.05×10^-3 mm^2/s、0.55、0.53,诊断的敏感度分别为75.0％、65.0％、60.0％,特异度分别为100.0％、95.7％、97.8％,阳性预测值分别为100.0％、86.7％、92.3％,阴性预测值分别为90.2％、86.3％、84.9％,诊断符合率分别为92.4％、86.4％、86.4％.b=1000 s/mm^2时,良、恶性结节平均ADC值、rADC1000-1及rADC1000-2分别为（1.45±0.28）×10^-3、（0.93±0.08）×10^-3mm^2/s和0.75±0.16、0.53±0.09,0.74±0.15、0.52±0.07;其差异有统计学意义（t值分为11.844、5.820、8.082；P值均＜0.01）；ADC值、rADC1ooo-1及rADC1000-2阈值点分别为0.97×10^-3 mm^2/s、0.58、0.55,诊断的敏感度均为70.0％,特异度分别为100.0％、95.7％、93.5％,阳性预测值分别为100.0％、87.5％、82.4％,阴性预测值分别为88.5％、88.0％、87.8％,诊断符合率分别为90.9％、87.9％、86.5％.以上6种诊断方法的敏感度及诊断符合率差异无统计学意义（x2值分别为1.232、2.263,P值分别为0.942、0.812）.结论 ADC值与rADC值均为鉴别良恶性乳腺病变的重要参数,尤其是b= 800 s/mm^2时的ADC值临床应用价值最高.
Objective To estimate the applications of ADC value and rADC value in the diagnosis of nodular lesions of breasts.Methods Fifty-two cases with 66 nodular lesions of breasts confirmed by histopathology underwent diffusion-weighted magnetic resonance imaging.Three b values （0,800 and 1000 s/mm^2） were applied.The mean ADC values of the breast nodules,the ADC values of ipsilateral breast（ rADC1 ） and ADC values of contralateral breast （rADC2 ）were respectively measured.The independent-samples t-test and chi-square test were used for statistical analyses.Results Of the 52 patients,there were 18 patients with infiltrating ductal carcinoma and 34 patients with fibroadenoma.50 patients with 64 lesions were examined by DWI.（ 1 ） at b = 800 s/mm^2,the mean ADC values of malignant nodules [ （ 1.01 ±0.09） × 10^-3 mm^2/s],rADC800-1 （0.52 ±0.07）and rADC800-2 （0.51 ±0.06） were lower than that of the benign nodules [ ADC value = （ 1.54 ± 0.28 ） × 10 -3 mm^2/s,t = 8.217,P 〈 0.01 ; rADC800-1 =0.77 ±0.15,t =9.339,P〈0.01 ; rADC800-2 =0.76 ±0.14,t = 10.394,P 〈0.01 ].The one-side upper limits of 95% medical reference value of mean values of infiltrating ductal carcinoma were adopted as the threshold point to distinguish the malignant from the benign.The threshold value of breast malignant nodule ADC,the rADC800-1 and rADC800-2 were respectively 1.05 × 10^-3 mm^2/s,0.55 and 0.53.The sensitivities of the three methods were 75.0%,65.0% and 60.0% ; the specificities were 100.0%,95.7% and 97.8% ;the positive predictive values were respectively 100.0%,86.7% and 92.3% ; the negative predictive values were 90.2%,86.3% and 84.9%; the diagnosis accordance rates were respectively 92.4%,86.4% and 86.4%.（ 2 ） at b = 1000 s/mm^2,the mean ADC values of malignant nodules [ （ 0.93 ± 0.08 ） ×10^-3 mm^2/s],rADC1000-1 （0.53 ±0.09） and rADC1000-2 （0.52 ±0.07） were also lower than that of the benign nodules[ ADC value= （1.45 ±0.28） ×10^-3 mm^2/s,t=11.844,P〈0.01; rADC1000-1 =0.75 ±0.16,t=5.820,P 〈 0.01 ; rADC1000-2 = 0.74 ± 0.15,t = 8.082,P 〈 0.01 ].The threshold value points breast malignant nodule ADC,the rADC1000-1 and rADC1000-2 were respectively 0.97 × 10^-3 mm^2/s,0.58,0.55.The sensitivities were all 70.0% ; the specificities were respectively 100.0%,95.7% and 93.5% ;the positive predictive values were 100.0%,87.5% and 82.4% ; the negative predictive values were 88.5%,88.0% and 87.8% ; the diagnosis accordance rates were 90.9%,87.9% and 86.5% respectively.There were no significant differences in specificities and the diagnosis accordance rates （ x^2 = 1.232,2.263 ; P =0.942,0.812 ）.Conclusions ADC value and rADC value are both important parameters of MRI in differentiating benign and malignant breast diseases.The study indicated that ADC value （ at b =800 s/mm^2） was the most valuable parameter.
出处
《中华放射学杂志》
CAS
CSCD
北大核心
2011年第12期1117-1121,共5页
Chinese Journal of Radiology
关键词
乳腺疾病
磁共振成像
弥散
Breast diseases
Diffusion magnetic resonance imaging
