高场磁共振三维动态增强结合扩散加权成像对乳腺良性和恶性病变的鉴别诊断

摘要

目的：探讨三维动态增强结合磁共振扩散加权成像（DWI）对乳腺良、恶性病变诊断与鉴别诊断。方法回顾性分析经病理证实乳腺占位性病变46例患者，均行DWI和动态增强磁共振成像（MRI）扫描，按扩散敏感系数（b值）800及1000 s/mm2各2次扫描数据做出各自表观扩散系数（ADC）图像。分别总结依据ADC值和时间-信号强度曲线评价乳腺良、恶性病变的统计学意义及其敏感性、特异性、阳性预测值、阴性预测值，统计综合考虑ADC值和时间-信号强度曲线类型评价乳腺良、恶性病变。结果46例乳腺肿瘤中恶性病变28例，良性病变18例。当b=800 s/mm2时，良、恶性病变及正常乳腺组织ADC值分别（1.54±0.28）×10-3 mm2/s、（1.01±0.09）×10-3 mm2/s、（1.49±0.06）×10-3 mm2/s；当b=1000 s/mm2时，良、恶性病变及正常乳腺组织ADC值分别（1.45±0.28）×10-3 mm2/s、（0.90±0.08）×10-3 mm2/s、（1.49±0.09）×10-3 mm2/s；良性与恶性及正常与恶性两者之间差异有统计学意义（P值<0.05）。在b值800×10-3 s/mm2、1000×10-3 s/mm2时,恶性病变ADC值的95%参考值范围分别为（0.97~1.05）×10-3 mm2/s、（0.90~0.97）×10-3 mm2/s，将ADC阈值定为1.050、0.969时,诊断乳腺恶性病变的敏感度分别为75%、70%，特异度100%、100%。时间-信号曲线：28例恶性病变中流出型（Ⅲ型）18例（64%），平台型（Ⅱ型）9例（32%），流入型（Ⅰ型）1例（3.5%）；18例良性病变中流入型（Ⅰ型）13例（72%），平台型（Ⅱ型）3例（17%），流出型（Ⅲ型）1例（6%）；时间-信号曲线诊断标准判断其敏感度96%(27/28),特异度83%（15/18）。结论单独采用DWI或三维动态增强在乳腺疾病诊断方面均存在局限性，联合运用二者，可互补彼此不足之处，提高对乳腺良、恶性肿瘤的鉴别能力。%Objective Research the distinguish diagnosis of benign and malignant breast diseases with three-dimensional (3D) dynamic enhancement and DWI in high magnetic resonance imaging. Methods Forty-six patients who were suffered breast diseases confirmed by pathology had been scanned with DWI and 3D dynamic enhancement in breast magnetic resonance imaging (MRI), and the ADC pictures respectively were drawn according to different b values of 800 and 1 000 s/mm2. Statistically assess the two kinds of breast diseases′sensitivity, specificity, positive predictive value, and negative predictive value by time-signal intensity curve. Results Among 46 cases, there were 28 patients of malignant breast diseases including 24 cases of infiltrating ductal carcinoma, 3 cases of medullary car-cinoma, and 1 case of lobular carcinoma; 18 cases of benign breast diseases (10 cases of fibroadenoma, 2 cases of atypical hyperplasia, 4 cases of cystic hyperplasia, 2 cases of abscess). When b value was 800 s/mm 2, benign, malig-nant and normal breast tissue′s ADC value were separately (1.54±0.28)×10-3 mm2/s ,(1.01±0.09)×10-3 mm2/s, and (1.49 ±0.06) ×10-3 mm2/s; when it was 1 000 s/mm2, their values were (1.45 ±0.28) ×10-3 mm2/s, (0.90 ±0.08) ×10-3 mm2/s, and(1.49±0.09)×10-3 mm2/s. There was significant statistical difference (P<0.05). The 95%of the reference in-terval in ADC values of malignant and normal breast tissues were(0.97-1.05)×10-3 mm2/s, (0.90-0.97)×10-3 mm2/s. In 28 malignant breast diseases, there were 18 cases of inflow type (64%), 9 cases of platform type(32%), and 1 case of outflow type (3.5%); yet in 18 benign diseases, there were 13 cases of inflow type (72%), 3 cases of platform type (17%), and 1 case of outflow type (6%). The sensitivity, specificity, positive predictive value, and negative predictive value by time-signal intensity curve were 96%(27/28), 83%(15/18) respectively. Conclusion The application of combining 3D dynamic enhancement with DWI sequences could improve distinguish diagnosis of benign and malig-nant breast diseases.