动态增强MRI对乳腺FFDM检出的BI-RADS 4-5类非肿块型微钙化病灶的诊断价值

doi:10.3969/j.issn.1006-6233.2020.06.024

摘要
图/表
参考文献
相关文章 (4)

全文: PDF (1707 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要目的:探讨动态增强MRI(dynamic contrast enhanced magnetic resonance imaging,DCE-MRI)对全数字化乳腺摄影(full-field digital mammography,FFDM)检出的BI-RADS 4-5类非肿块型微钙化病灶的诊断价值。方法:回顾性分析本院2017年1月到2019年8月乳腺FFDM检出的BI-RADS 4-5类非肿块型微钙化病灶、均行双侧乳腺DCE-MRI检查、皆有病理结果的148例患者资料,共152个病灶。绘制两种诊断方法的接受者工作特征曲线(receiver operator characteristic curve,ROC曲线),根据最大约登指数寻找最佳诊断阈值,作为判断两种诊断方法对乳腺病灶BI-RADS分类的良恶性界限;以病理结果为金标准,计算两种检查方法诊断对不同乳腺X线摄影BI-RADS分类病变的敏感度、特异度、阳性预测值(positive predictive valve,PPV)、阴性预测值(negative predictive value,NPV)、准确度,采用χ²检验或Fisher精确检验比较两种诊断方法对不同BI-RADS分类的敏感度、特异度、PPV、NPV、准确度,并采用Z检验比较两种诊断方法的ROC曲线下面积(area under curve,AUC)。结果:148例患者一共152个病灶,经病理活检证实为良性62个,恶性90个,绘制两种诊断方法的ROC曲线,得出最佳诊断阈值均为4b,即判断两种诊断方法的BI-RADS 4b以下为良性,BI-RADS 4b及其以上为恶性;对于乳腺FFDM检出乳腺BI-RADS 4-5类非肿块型微钙化总共152病灶,FFDM诊断病灶的敏感度、特异度、PPV、NPV、准确度分别为86.7%、53.2%、72.9%、73.3%、73.0%,DCE-MRI诊断病灶的敏感度、特异度、PPV、NPV、准确度分别为93.3%、83.9%、89.4%、89.7%、89.5%,两者的敏感度之间差异无统计学意义(χ²值为2.222,P值为0.136>0.05),但两者的特异度、PPV、NPV、准确度之间差异均有统计学意义(χ²值分别为13.503、8.674、4.681、13.495,P值均<0.05);对于FFDM检出乳腺BI-RADS 4类非肿块型微钙化总共125个病灶,FFDM诊断病灶的敏感度、特异度、PPV、NPV、准确度分别为82.4%、57.9%、70.0%、73.3%、71.2%,DCE-MRI诊断病灶的敏感度、特异度、PPV、NPV、准确度分别为92.6%、86.0%、88.7%、90.7%、89.6%,两者的敏感度之间差异无统计学意义(χ²值为3.294,P值为0.070>0.05),但两者的特异度、PPV、NPV、准确度之间差异均有统计学意义(χ²值分别为11.122、7.903、5.229、13.428,P值均<0.05);对于FFDM检出乳腺BI-RADS 5类非肿块型微钙化总共27个病灶,FFDM诊断病灶的敏感度、特异度、PPV、NPV、准确度分别为100.0%、0、81.5%、0、81.5%,DCE-MRI诊断病灶的敏感度、特异度、PPV、NPV、准确度分别为95.5%、60.0%、91.3%、75.0%、88.9%,两者的敏感度、特异度、PPV、准确度之间差异均无统计学意义(P值均>0.05);对于乳腺FFDM检出乳腺BI-RADS 4-5类非肿块型微钙化,FFDM及DCE-MRI诊断方法的ROC曲线下面积(AUC)分别为0.752、0.907,两者AUC之间差异有显著统计学意义(Z值为4.068,P<0.0001,95%置信区间CI为0.0804~0.230)。结论:对于FFDM检出乳腺BI-RADS 4类非肿块型微钙化病灶,DCE-MRI能有效提高病变的诊断效能;对于FFDM检出乳腺BI-RADS 5类非肿块型微钙化病灶的定性诊断,DCE-MRI无明显优势。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章

关键词 ：全数字化乳腺摄影, 动态对比增强, 磁共振成像, 非肿块型微钙化

Abstract：Objective:To investigate the diagnostic value of dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) in the diagnosis of BI-RADS 4-5 non-mass microcalcification lesions on full-field digital mammography (FFDM). Methods: Conduct retrospective analysis of 148 patients with 152 lesions, who were examined by bilateral breast DCE-MRI and all had pathological results,from January 2017 to August 2019. Plot the receiver operator characteristic curve (ROC) of the two diagnostic methods, and verify the best diagnostic threshold according to the most approximate index, which was used to judge the benign and malignant boundary of the BI-RADS classification of breast lesions. With histopathological results as the golden standard, the sensitivity, specificity and positive predictive value (PPV), negative predictive value (NPV) and accuracy of the two methods for different BI-RADS classification were calculated and compared with χ² test or Fisher exact test, and the area under the ROC curve (AUC) of the two diagnostic methods were compared with Z test. Results: There were 152 lesions in 148 patients including 62 benign lesions and 90 malignant lesions confirmed by pathological biopsy. The ROC curves of the two diagnostic methods were plotted, and the best diagnostic threshold was 4b, the BI-RADS below 4b was benign and the BI-RADS 4b or above was malignant of the two diagnostic methods. For 152 cases of breast BI-RADS 4-5 non-mass microcalcification lesions on FFDM, the sensitivity, specificity, PPV, NPV and accuracy of FFDM were 86.7%, 53.2%, 72.9%, 73.3%, 73.0%, DCE-MRI was 93.3%, 83.9%, 89.4%, 89.7%, 89.5%, respectively. The sensitivity of the two methods was not significantly different (χ²=2.222, P>0.05), and the specificity, PPV, NPV and accuracy of the two methods were statistically significantly different (χ²=13.503, 8.674, 4.681, 13.495,respectively, all P<0.05). For 125 cases of breast BI-RADS 4 non-mass microcalcification lesions on FFDM, the sensitivity, specificity, PPV, NPV and accuracy of FFDM were 82.4%, 57.9%, 70.0%, 73.3%, 71.2%, DCE-MRI was 92.6%, 86.0%, 88.7%, 90.7%, 89.6%, respectively. The sensitivity of the two methods was not significantly different (χ²=3.294, P>0.05), the specificity, PPV, NPV and accuracy of the two methods were statistically significantly different (χ²=11.122, 7.903, 5.299, 13.428, respectively, all P<0.05). For 27 cases of breast BI-RADS 5 non-mass microcalcification lesions on FFDM, the sensitivity, specificity, PPV, NPV and accuracy of FFDM were 100.0%, 0, 81.5%, 0, 81.5%, DCE-MRI was 95.5%, 60.0%, 91.3%, 75.0%, 88.9%, respectively. The sensitivity specificity, PPV and accuracy of the two methods were not significantly different (all P＞0.05). For breast BI-RADS 4-5 non-mass microcalcification lesions on FFDM, the areas under the ROC curve (AUC) of FFDM and DCE-MRI diagnostic methods were 0.752 and 0.907, respectively. The difference between the two AUC was statistically significant. (Z=4.068, P＜0.0001, 95% confidence interval CI was 0.0804~0.230). Conclusion: For breast BI-RADS 4-5 non-mass microcalcification lesions on FFDM, DCE-MRI can effectively improve the diagnostic efficiency of lesions; for breast BI-RADS 4-5 non-mass microcalcification lesions on FFDM, DCE-MRI has no obvious advantage.

Key words： Full digital mammography Dynamic contrast enhancement Magnetic resonance imaging Non-mass microcalcification

通讯作者: 刘庆祥

引用本文:

解丽丽, 白仁静, 郑义, 刘庆祥. 动态增强MRI对乳腺FFDM检出的BI-RADS 4-5类非肿块型微钙化病灶的诊断价值[J]. 河北医学, 2020, 26(6): 988-993.
XIE Lili, BAI Renjing, et al. Diagnostic Value of Dynamic Contrast Enhanced MRI in Examination of BI-RADS 4-5 Non-mass Microcalcification Lesions on FFDM. HeBei Med, 2020, 26(6): 988-993.

链接本文:

http://www.hbyxzzs.cn/CN/10.3969/j.issn.1006-6233.2020.06.024 或 http://www.hbyxzzs.cn/CN/Y2020/V26/I6/988

[1] Bluekens AM, Holland R, Karssemeijer N, et al. Comparison of digital screening mammography and screen-film mammography in the early detection of clinically relevant cancers: a multicenter study [J]. Radiology,2012, 265(3): 707~714.
[2] D Orsi CJ, Bassett LW, Appleton CM, et al. Breast imaging reporting and data system: ACR BI-RADS-breast imaging atlas[M]. Reston, VA: American College of Radiology, 2013.
[3] 盛蕾,锁彤,张霞,等.对比增强乳腺X线摄影对于致密型乳腺乳腺癌的诊断价值[J].中华放射学杂志,2019,53(2):98~102.
[4] 师金,梁迪,李道娟,等.全球女性乳腺癌流行情况研究[J].中国肿瘤,2017,26(9):683~690.
[5] Rominger MB, Sax EV, Figiel JH, et al. Occurrence and positive predictive value of additional nonmass findings for risk stratification of breast microcalcifications in mammography [J]. Can Assoc Radiol J, 2013, 64(4): 333~338.
[6] Meagher T, Trop I. Breast Microcalcifications and their implications for risk assessment. A brief review [J].Insur Med, 2017, 47(1): 40~42.
[7] de Leeuw H, Stehouwer BL, Bakker CJ, et al.Detecting breast microcalcifications with high-field MRI [J].NMR Biomed, 2014, 27(5): 539~546.
[8] Ha SM, Cha JH, Shin HJ, et al. Mammography, US, and MRI to assess outcomes of invasive breast cancer with extensive intraductal component: a matched cohort study [J]. Radiology, 2019, 292(2): 299~308.
[9] 李二妮,李静,宋颖,等.MRI对乳腺X线摄影检出的微钙化病变诊断的价值[J].中华放射学杂志,2014,48(12):1005~1008.
[10] 陈佳儿,肖莹,江红,等.动态增强MRI在乳腺X线摄影检出BI-RADS4~5级微钙化病变的价值[J].临床放射学杂志,2018,37(2):222~226.
[11] Li E, Li J, Song Y, et al. A comparative study of the diagnostic value of contrast-enhanced breast MR imaging and mammography on patients with BI-RADS 3-5 microcalcifications [J]. PLoS One, 2014, 9(11): e111217.
[12] Eun NL, Son EJ, Gweon HM, et al. The value of breast MRI for BI-RADS category 4B mammographic microcalcification: based on the 5(th) edition of BI-RADS [J]. Clin Radiol, 2018, 73(8): 750~755.
[13] Bennani-Baiti B, Dietzel M, Baltzer PA. MRI for the assessment of malignancy in BI-RADS 4 mammographic microcalcifications [J]. PLoS One,2017,12(11):e0188679.