摘要
目的本研究的目的是荟萃分析乳腺核磁共振成像(MRI)的各个恶性肿瘤征象,为MR诊断及鉴别诊断乳腺肿瘤性质提供依据。方法文献检索查找乳腺核磁共振成像的各个恶性征象提示,并经由病理确诊的乳腺恶性肿瘤的试验,采用Meta-Disc 1.4软件,对各个MRI恶性征象的诊断效能进行逐一评价。结果按照文献纳入及排除标准最终有19项独立研究纳入本次Meta分析。Meta分析结果显示:1毛刺状形态AUC值为0.8167,Q值0.7506。2边缘模糊不清AUC值为0.6881,Q值0.6435。3环状强化AUC值为0.6879,Q值0.6433。4washout曲线AUC值为0.7955,Q值0.7320。5ADC值测量AUC值为0.5610,Q值0.5459。采用随机效应模型合并MRI诊断乳腺恶性肿瘤的总灵敏度为0.497,特异性为0.810,AUC值为0.7071,Q值0.6585。结论乳腺核磁共振成像(MRI)的各个恶性肿瘤征象对于恶性肿瘤的价值效能不同,恶性征象中毛刺状形态和动态增强的时间-信号曲线的washout型曲线对乳腺恶性肿瘤的诊断价值较高,而环状强化和ADC值的测量的诊断价值较差。
Objective The purpose of this study is to meta-analysis the diagnostic value of breast magnetic resonance imaging （MILl） of various malignancies signs. Methods Using Meta-Disc 1.4 software, diagnostic effectiveness of every MILl diagnosis signs was evaluated one by one. ,Results In accordance with the inclusion and exclusion criteria, 19 independent studies eventually included in this Meta-analysis. Meta-analysis showed that： burr-like shape AUC value of 0.8167, Q value of 0.7506. blurred edges AUC value of 0.6881, Q value of 0.6435. ring enhancement AUC value of 0.6879, Q value of 0.6433. @washout curve AUC value of 0.7955, Q value of 0.7320. @ADC measurement AUC value of 0.5610, Q value of 0.5459. Using a random effects model merge MI〈I diagnosis of breast cancer overall sensitivity was 0.497, specificity was 0.81（）, AUC value of 0.7071, Q value of 0.6585. CondtLrdon Various malignancies signs of breast magnetic resonance imaging （MRI） provide different effectiveness for diagnosis, burr-like morphology and dynamic contrast-enhanced washout curve signs have higher diagnostic predictive value for breast cancer, while the diagnostic value of measuring ring enhancement and ADC values are poor.
出处
《中国CT和MRI杂志》
2015年第1期1-4,10,共5页
Chinese Journal of CT and MRI
