扩散加权成像与传统磁共振成像联合诊断乳腺疾病

发布时间：2018-01-21 01:17

本文关键词： 乳腺肿瘤磁共振成像扩散加权成像表观扩散系数病理学外科诊断鉴别　出处：《中国医学影像学杂志》2016年04期 　论文类型：期刊论文

【摘要】：目的探讨扩散加权成像(DWI)与常规MRI征象联合诊断乳腺疾病的模式及诊断效能。资料与方法采用前瞻性研究,99例乳腺疾病患者共109个病变术前行3.0T乳腺MRI及DWI检查,经术后病理或长期随访确诊。乳腺良恶性病变的表观扩散系数(ADC)值进行同一观察者、不同观察者共3次测量,采用受试者操作特征曲线、串联并联诊断试验、机器学习探讨多参数联合对乳腺良恶性疾病的诊断效能。结果 ADC值的观察者内组内相关系数为0.950(95%CI 0.928~0.965);观察者间组内相关系数为0.937(95%CI 0.909~0.956);良性病变与恶性病变的ADC值分别为(1.447±0.377)×10-3 mm2/s、(1.054±0.255)×10-3 mm2/s,差异有统计学意义(P0.05)。以ADC值1.125×10-3 mm2/s为阈值鉴别良恶性病变的敏感度和特异度分别为82.4%、74.1%,联合其他传统MRI征象可使敏感度、特异度高于单一ADC值诊断。结论 ADC值具有良好的可重复性和稳定性,ADC值与传统MRI征象联合应用可以更好地鉴别乳腺病灶的良恶性。
[Abstract]:Objective to explore the mode and diagnostic efficacy of DWI combined with conventional MRI in the diagnosis of breast diseases. 99 patients with breast disease underwent 3.0T MRI and DWI examination before operation for 109 lesions. The apparent diffusion coefficient (ADCC) of benign and malignant breast lesions was measured for the same observer and different observers for 3 times. Series and parallel diagnostic test. Machine learning was used to investigate the effectiveness of multiparameter combination in the diagnosis of breast benign and malignant diseases. Results the intraobserver correlation coefficient of ADC was 0.950 卤95 CI 0.928 0.965. The correlation coefficient in the inter-observer group was 0.93795 CI 0.909 ~ 0.956; The ADC values of benign and malignant lesions were 1.447 卤0.377 脳 10 ~ (-3) mm ~ (-2) / s ~ (-1) 卤(1.054 卤0.255) 脳 10 ~ (-3) mm2/s, respectively. The sensitivity and specificity of ADC 1.125 脳 10 ~ (-3) mm2/s for differentiating benign and malignant lesions were 82.4% and 74.1%, respectively. Combined with other traditional MRI signs, sensitivity and specificity were higher than that of single ADC. Conclusion the ADC value has good reproducibility and stability. The combination of ADC value and traditional MRI signs can better distinguish benign and malignant breast lesions.
【作者单位】：解放军第四军医大学唐都医院放射科;西北大学信息科学与技术学院;
【基金】：陕西省科技攻关项目(2015SF119) 第四军医大学青年英才支持计划(第一层次) 第四军医大学唐都医院科技创新发展基金自主创新探索项目(2013CXTS005);第四军医大学唐都医院科技创新发展基金临床研究重大项目(2015LCYJ001)
【分类号】：R445.2;R737.9
【正文快照】： Affiliated to the Fourth Military MedicalUniversity,Xi'an 710038,ChinaAddress Correspondence to:CHEN BaoyingE-mail:chenby128@163.com床研究重大项目(2015LCYJ001)。修回日期:2016-02-20中国医学影像学杂志2016年第24卷第4期:270-276Chinese Journal of Medical Imag

【相似文献】