MRCP联合高分辨率DWI对非结石性肝外胆管梗阻性病变定性诊断的初步研究

发布时间：2018-04-23 14:08

本文选题：胆道梗阻 + 弥散加权成像　；参考：《大连医科大学》2014年硕士论文

【摘要】：目的：探讨胰胆管水成像（MRCP）联合高分辨率弥散加权成像(HR-DWI)对非结石性肝外胆道梗阻性病变定性诊断的诊断价值。材料与方法：回顾性分析2013年8月至2014年2月于我院行上腹部MR检查的资料完整的43例非结石性肝外胆道梗阻病例，其中恶性肝外胆道梗阻病例25例，包括胆总管癌18例，壶腹癌3例，十二指肠乳头癌4例；良性肝外胆道梗阻18例，包括胆管炎15例、十二指肠乳头炎症3例；男：女=19：24；年龄61.53±12.54岁。扫描序列包括：轴位T2WI、MRCP及胆道梗阻端连续多层轴位STD-DWI序列（矩阵128×128、FOV=44x44cm2，S组）和轴位薄层HR-DWI序列（矩阵196×96、FOV=31x15.5cm2，H组），b值均为600s/mm2。两名腹部影像诊断医师在双盲情况下进行图像分析。主观分析包括：①根据胆管梗阻的MRCP影像学特征，参照Baron等诊断标准评判其对良恶性病变的诊断符合率；②两名观察者分别对STD-DWI序列（S组）、HR-DWI序列（H组）图像质量及病变的显示情况进4级评分，并采用Kappa一致性检验来评价两名观察者评分的一致性；对S、H两组评分使用卡方检验组间评分统计学差异；③两名观察者共同对单独MRCP序列（A组）、MRCP联合T2W（IB组）、MRCP联合STD-DW（IC组）、MRCP联合HR-DWI（D组）对病变良恶性进行3级评分判定，对四组评分分别使用卡方检验进行组间统计学分析，后使用ROC曲线分析四组对恶性梗阻病变诊断的AUC值、准确率、敏感度、特异度、阳性预测值、阴性预测值。客观评价包括：①计算两组DWI序列的图像分辨率；②两名观察者分别测量S、H两组梗阻端病变（a）、邻近正常胆管壁（b）的ADC值，应用ICC检验两名观察者测量值的一致性，对S、H两组所测a、b两处的ADC值进行Wilcoxon秩和检验；采用ROC曲线分析H组a、b两处所测得的ADC值对鉴别胆道梗阻端病变的良恶性的诊断效能。结果：①MRCP良性梗阻征象13例，其诊断良性符合率（7/13，53.85%)；MRCP恶性梗阻征象30例，其诊断恶性符合率（19/30，63.33%）。②两名观察者对STD-DWI的病灶显示评分分别为（3.20±0.56vs.3.23±0.61），值0.867；两名观察者对HR-DWI的图像质量评分分别为（3.81±0.39vs.3.79±0.41），，值0.927；S、H两组评分卡方检验（Χ2=15.878，p=0.009）。③两名观察者共同对A、B、C、D四组扫描方案对肝外胆道梗阻端病灶定性进行3级评分，四组组间评分卡方检验结果分别为P值0.05。③两名观察者共同对A、B、C、D四组扫描方案对肝外胆道梗阻端病灶定性进行3级评分，评分结果分别为：2.49±0.51、2.47±0.67、2.47±0.67、2.42±0.82，AUC值分别为：0.681、0.811、0.854、0.944，准确率分别为：67.44%、79.07%、83.72%、97.62%，敏感度分别为：64.00%、80.00%、84.00%、100%，特异度分别为：72%、77.8%、83.3%、88.9%，阳性预测值分别为：76.19%、83.33%、87.50%、92.59%，阴性预测值分别为：59.09%、73.68%、78.95%、100%；④S、H组的空间分辨率分别约为3.4，1.6（cm/pixel）；⑤S组a、b部位两位观察者测量的ADC值（×10-3mm2/s）为：（1.81±0.51，2.02±0.46）vs.（1.87±0.61，2.04±0.47），H组a、b部位两位观察者测量的ADC值（×10-3mm2/s）为：（1.78±1.28，1.79±1.15）vs.（1.78±1.33，1.78±1.19），两名观察者所测的数据ICC值分别为0.996、0.998、0.970、0.943。S组良恶性梗阻部位a、b两处的ADC值间差异无统计学意义（Z=-1.330，-1.404；P值0.184，0.160），H组良恶性梗阻部位a、b两处的ADC值间差异有统计学意义（Z=-2.327，-2.413；P值0.020，0.016）。H组a、b部位ADC值诊断恶性胆道梗阻的AUC值为（0.710：0.718）。结论：HR-DWI较STD-DWI图像分辨率高，对肝外胆道梗阻端病变显示更为清楚；MRCP联合HR-DWI序列大大提高对肝外胆管恶性病变的诊断效能；HR-DWI的ADC值对良恶性病变的鉴别有一定帮助。
[Abstract]:Objective: To evaluate the diagnostic value of cholangiopancreatography (MRCP) combined with high resolution diffusion weighted imaging (HR-DWI) for the diagnosis of non calculous extrahepatic biliary obstruction.
Materials and methods: a retrospective analysis of 43 cases of non calculous extrahepatic biliary obstruction from August 2013 to February 2014 in our hospital, including 25 cases of malignant extrahepatic biliary obstruction, including 18 cases of choledochal carcinoma, 3 ampullary carcinoma, 4 duodenal papilla carcinoma, 18 benign extrahepatic biliary obstruction, including bile duct, including bile duct, 43 cases of non calculous extrahepatic biliary obstruction. 15 cases of inflammation and 3 cases of duodenal papilla inflammation, male: female =19:24, age 61.53 + 12.54 years old. The scan sequence includes: axis position T2WI, MRCP and continuous multi-layer STD-DWI sequence of biliary obstruction end (matrix 128 x 128, FOV=44x44cm2, S group) and axial thin layer HR-DWI sequence (matrix 196 * 96, FOV=31x15.5cm2, H group), b values are 600s/mm2. two abdomen shadow Image analysis was performed under double blindness as a diagnostic physician. Subjective analysis included: (1) according to the MRCP imaging features of bile duct obstruction, the diagnostic coincidence rate of benign and malignant lesions was judged by Baron and other diagnostic criteria; (2) two observers were divided into 4 levels of STD-DWI sequence (group S), HR-DWI sequence (group H) image quality and pathological changes. Score, and use Kappa consistency test to evaluate the consistency of two observer scores; for S, group H two scores using chi square test group scores statistical differences; (3) two observers shared a single MRCP sequence (group A), MRCP combined T2W (IB group), MRCP joint STD-DW (IC group), MRCP combined with 3 grades for benign and malignant lesions. The four groups were judged by chi square test for statistical analysis, and then the AUC value, accuracy, sensitivity, specificity, positive predictive value and negative predictive value were analyzed by the ROC curve in the four groups. The objective evaluation included: (1) the resolution of the image of two groups of DWI sequences was calculated; and two observers measured respectively. S, H two groups of obstructive end lesions (a), adjacent normal bile duct wall (b) ADC value, ICC test of the consistency of the measured values of two observers, S, H two groups measured a, B two ADC value at the Wilcoxon rank and test, two measured values for the differential diagnosis of biliary obstruction end lesions of the benign and malignant diagnostic effectiveness.
Results: (1) 13 cases of benign obstruction of MRCP, the benign coincidence rate of diagnosis (7/13,53.85%), the signs of malignant MRCP obstruction in 30 cases, and the malignant coincidence rate (19/30,63.33%). (2) the lesions of two observers were (3.20 + 0.56vs.3.23 + 0.61) and 0.867, and two observers were respectively (3.) of the image quality of HR-DWI, respectively. 81 + 0.39vs.3.79 + 0.41), value 0.927, S, H two score chi square test (2=15.878, p=0.009). (3) two groups of observers shared a grade 3 score on A, B, C, and D four groups on the extrahepatic biliary obstruction end focus, and the scores of the four groups were P values 0.05., two of the two observers, four groups of scanning schemes for the liver The 3 grade score of the external biliary obstruction end focus was made. The results were 2.49 + 0.51,2.47 + 0.67,2.47 + 0.67,2.42 + 0.82 respectively, and the AUC values were 0.681,0.811,0.854,0.944, the accuracy was 67.44%, 79.07%, 83.72%, 97.62%, respectively: 64%, 80%, 84%, 100%, respectively: 72%, 77.8%, 77.8%, 83.3%, Yang, Yang, respectively The predictive values of the sex were 76.19%, 83.33%, 87.50%, 92.59%, and the negative predictive values were 59.09%, 73.68%, 78.95%, 100%; the spatial resolution of group H was 3.4,1.6 (cm/pixel), S group A and ADC value (1.81 + 0.51,2.02 + 0.46) vs. (1.87 + 0.61,2.04 0.47) in B site two observers (1.81 + 0.51,2.02 + 0.46) The ADC value (x 10-3mm2/s) measured by the observer was (1.78 + 1.28,1.79 + 1.15) vs. (1.78 + 1.33,1.78 + 1.19). The ICC values measured by two observers were a of the benign and malignant obstructive sites in the 0.996,0.998,0.970,0.943.S group, and there was no statistical difference between the ADC values of the B two (Z =-1.330). There was a significant difference in the ADC values between the two sites (Z=-2.327, -2.413, P value 0.020,0.016), a in.H group, and ADC value in B position for diagnosing malignant biliary obstruction (0.710:0.718).
Conclusion: the resolution of HR-DWI is higher than that of STD-DWI image, and it is more clear to the extrahepatic biliary obstruction end lesions; MRCP combined with HR-DWI sequence can greatly improve the diagnostic efficiency of extrahepatic bile duct malignant lesions, and the ADC value of HR-DWI has some help for the identification of benign and malignant lesions.

【学位授予单位】：大连医科大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R445.2;R657.4

【参考文献】