MRI多序列成像在膀胱癌分期及分级中的应用研究

发布时间：2018-03-12 15:38

本文选题：扩散加权成像　切入点：磁共振成像　出处：《华中科技大学》2016年博士论文　论文类型：学位论文

【摘要】：第一部分小视野及常规视野DWI在非肌层及肌层侵袭性膀胱癌中的应用目的：比较小视野弥散加权成像(reduced field-of-view diffusion-weighted imaging, rFOV DWI)与常规视野弥散加权成像(full field-of-view diffusion-weighted imaging, fFOV DWI)在肌层及非肌层侵袭性膀胱癌中的图像质量、诊断准确性及表观弥散系数(apparent diffusion coefficient, ADC)的应用价值。材料与方法：经华中科技大学同济医学院附属同济医院伦理委员会批准,39例膀胱癌病人共60个肿瘤病灶分别行rFOV DWI、fFOV DWI及常规MRI成像。所有肿瘤均得到了病理学分期及组织学分级的结果。两位放射学诊断学者对两种DWI图的质量采用四分法分别进行评分。另外,两名腹部放射学诊断医生分别对三组影像图像(T2WI, T2WI+fFOV DWI及T2WI+rFOV DWI)在不知道病理结果的情况下分别进行膀胱癌分期的诊断,并对膀胱癌分期为T2或以上的诊断信心进行评分。两种DWI图像质量的评估采用维氏符号秩次检验：每组图像的诊断准确性、敏感性及特异性的比较采用McNemar检验；诊断效能用受试者工作曲线(receiver operating characteristic curves, ROC)下面积表示；用Mann-Whitney U检验来比较不同膀胱癌分期及分级的ADC值的差别。结果：rFOV DWI的图像质量(平均3.62)明显高于fFOV DWI的(平均2.98,p 0.001)。T2WI, T2WI+ fFOV DWI及T2WI+rFOV DWI对膀胱癌分期的诊断准确性分别为57%、70%和78%。加rFOV DWI后诊断T2及以上分期的准确性及特异度的诊断效能明显提高了(p0.05)。无论是rFOV DWI还是fFOV DWI,肌层侵袭性膀胱癌及高级别膀胱癌的ADC值均明显低于非肌层侵袭性及低级别膀胱癌的ADC值(P0.01)。结论：rFOV DWI无论在图像质量上还是在诊断准确性上均优于fFOV DWI,在常规MRI检出病灶的基础上行r-FOV DWI,对评估肿瘤是否浸润肌层很有价值。两种DWI序列的ADC值可能有助于鉴别肌层侵袭性膀胱癌及非肌层侵袭性膀胱癌,可能有助于区别高级别膀胱癌及低级别膀胱癌。第二部分不同数学模型扩散加权成像在膀胱癌分期及分级中的应用研究目的：探讨不同数学模型(单指数模型、双指数模型及拉伸指数模型)扩散加权成像在膀胱癌分期及分级中的应用价值。材料与方法：经华中科技大学同济医学院附属同济医院伦理委员会批准,37例膀胱癌病人共51个肿瘤病灶分别行常规MRI及多b值DWI成像。所有肿瘤均得到了病理学分期及组织学分级的结果。两位放射学诊断学者在不知道病理结果的情况下分别对膀胱癌进行多b值DWI不同数学模型的数据测量。两位测量者间所测值一致性由Bland Altman来分析。用Mann-Whitney U检验来比较不同膀胱癌分期及分级的单指数模型量化值ADC、双指数模型量化值D(真性扩散系数,the true diffusioncoefficient)、D*(灌注相关的假性扩散系数,perfusion-related pseudo-diffusion coefficient)、f(灌注分数,Perfusion fraction)及拉伸指数模型量化值DDC(分布扩散系数,the distributed diffusion coefficient)和α(不均质性指数,heterogeneity index)的差别。另外,用受试者工作曲线(receiver operating characteristic, ROC)来判断每个量化指标在鉴别高低级别膀胱癌及肌层侵袭性和非肌层侵袭性膀胱癌中诊断效能。结果：两位测量者间的一致性较好。肌层侵袭性膀胱癌或高级别膀胱癌的ADC、D及DDC值明显低于非肌层侵袭性膀胱癌或低级别膀胱癌的相应值(P=-0.002,0.003和0.007；P=0.014,0.002和0.033)。而D*、f和α值在鉴别诊断膀胱癌的分期及组织学分级上却没有明显统计学差异。ADC、D及DDC值鉴别膀胱癌的分期的ROC曲线下面积(the areas under the receiver operating characteristic curves, Az)分别为0.751、0.747及0.726：在鉴别膀胱癌组织学分级上的Az分别为0.733、0.796和0.702；ADC、D及DDC值鉴别膀胱癌的分期及分级的Az两两比较均无统计学差异。结论：ADC、D和DDC值可以有效地鉴别膀胱癌是肌层侵袭性还是非肌层侵袭性,同时也可以有效地鉴别膀胱癌是高级别还是低级别；但是,它们的Az即诊断效能却没有明显的差异。三种指数模型的DWI成像均可以用来有效地对鉴别膀胱癌的分期及组织学分级。第三部分BOLD成像R2*值在膀胱癌分期及分级的初步应用研究目的：本研究的目的是探讨BOLD成像中R2*值在鉴别膀胱癌分期及分级中的应用价值。材料与方法：本研究经过了华中科技大学同济医学院附属同济医院伦理委员会批准,所有纳入研究的病人均签署了知情同意书。92例怀疑膀胱癌的病人进行了常规MRI扫描及T2*Mapping序列扫描,经过筛查最终纳入分析的病人数为57例。57例病人共90个膀胱癌病灶,所有膀胱癌病灶均获得了病理分期及组织学分级的结果。两位放射学诊断学者在不知道病理结果的情况下对膀胱癌进行R2*值的数据测量。两位测量者间所测值一致性由Bland Altman来分析。用Mann-Whitney U检验来比较不同膀胱癌分期及分级的R2*值的差别。另外,用受试者工作曲线(receiver operating characteristic, ROC)来判断每个量化指标在鉴别高低级别膀胱癌及肌层侵袭性和非肌层侵袭性膀胱癌中的诊断效能。结果：两位测量者间的一致性较好。高、低级别膀胱癌的R2*值分别为20.97±8.91Hz、16.43±5.74 Hz,二者具有明显统计学差异(P=-0.002)；肌层侵袭性及非肌层侵袭性膀胱癌的R2*值分别为21.35±8.96Hz、18.48±7.71Hz,二者比较具有统计学差异(P=-0.046)。R2*值鉴别诊断高低级别膀胱癌的ROC曲线下面积为0.703,鉴别肌层侵袭性膀胱癌及非肌层侵袭性膀胱癌的ROC曲线面积为0.636。结论：BOLD序列的R2*值可以很好地鉴别膀胱癌的分级及分期,可以为临床医生对膀胱癌的临床治疗方案的选择提供有利的信息。
[Abstract]:The first part of the small field of view and conventional view of DWI in non muscle and muscle invasive bladder cancer in application: comparison of small vision diffusion weighted imaging (reduced field-of-view diffusion-weighted imaging, rFOV DWI) and the conventional view of diffusion weighted imaging (full field-of-view diffusion-weighted imaging, fFOV DWI) invasive bladder cancer image quality in muscle layer and muscle layer, diagnostic accuracy and apparent diffusion coefficient (apparent diffusion, coefficient, ADC) application value. Materials and methods: the Tongji Medical College of Huazhong University of Science and Technology, affiliated Tongji Hospital Ethics Committee approval, 39 cases of bladder cancer patients with a total of 60 tumors were rFOV DWI, fFOV DWI and conventional MRI imaging. All the tumors were the histological pathological staging and histological results. Using four methods, two scholars of the two kinds of radiological diagnosis of DWI map quality respectively. Score of two. In addition, the diagnosis of abdominal radiology doctor respectively for the three groups of images (T2WI, T2WI+fFOV DWI and T2WI+rFOV DWI) do not know in the absence of histopathologic findings were staging diagnosis of bladder cancer, and the staging of bladder cancer is T2 or more diagnostic confidence for a score. Evaluation of two kinds of DWI image quality the Vivtorinox signed rank test: the diagnostic accuracy of each image, compared with the sensitivity and specificity of the McNemar test; diagnostic efficiency with receiver operating curve (receiver operating characteristic curves, ROC) area under representation; with Mann-Whitney U test to compare different bladder cancer staging and grading of ADC value. The difference between the results: the image quality of rFOV DWI (average 3.62) was significantly higher than that of fFOV DWI (average 2.98 P, 0.001.T2WI), T2WI+ fFOV DWI and T2WI+rFOV DWI on the staging of bladder cancer diagnosis accuracy. As of 57%, 70% and 78%. and rFOV DWI after the diagnosis efficiency and specificity of T2 and above the staging accuracy was significantly improved (P0.05). Both the rFOV DWI or fFOV DWI, muscle invasive bladder cancer and high-grade bladder cancer ADC values were significantly lower than those of non muscle invasive bladder and low level cancer ADC (P0.01). Conclusion: rFOV DWI in terms of image quality or diagnostic accuracy is superior to fFOV DWI, based on r-FOV DWI lesions were detected in conventional MRI, to assess whether tumor infiltrating muscular layer is of great value. The ADC value of two DWI sequences may contribute to bladder cancer and non myometrial invasion in differentiating muscle invasive bladder cancer, may contribute to the difference between high-grade bladder cancer and low grade bladder cancer. In the second part, different mathematical model of application of diffusion weighted imaging in bladder cancer Objective: To investigate the staging and grading of different mathematical models (single finger The number of model, double exponential and stretched exponential model) application value of diffusion weighted imaging in the staging and grading of bladder cancer. Materials and methods: approved by the ethics committee of Tongji Hospital Affiliated to Tongji Medical College of Huazhong University of Science and Technology, 37 cases of bladder cancer patients with a total of 51 lesions were examined with conventional MRI and b value of DWI imaging. All the tumors were the histological pathological staging and histological results. Two diagnostic radiology scholars in unaware of the pathological results of bladder cancer cases were multi B DWI values of different mathematical models of data measurement. Two measurement between the measured value to analyze the consistency by Bland Altman. Mann-Whitney U test to compare the different bladder cancer staging and grading of the single index model to quantify the value of quantitative ADC, double exponential model (true value of D diffusion coefficient, the true, D* (diffusioncoefficient) perfusion related pseudo diffusion coefficient, Perfusion-related pseudo-diffusion coefficient (f), Perfusion perfusion fraction, fraction) quantization and tensile index model DDC (the distributed diffusion distribution of diffusion coefficient, coefficient (alpha) and heterogeneity index, heterogeneity, index) the difference. In addition, receiver operating curve (receiver operating, characteristic, ROC) to determine the quantitative index of each invasive and non muscle invasive bladder cancer diagnosis in the differential diagnosis of bladder cancer and high level muscle. Results: the consistency between the two examiners. The better muscle invasive bladder cancer or high-grade bladder cancer ADC, D and DDC were significantly lower than the corresponding value of non muscle invasive bladder cancer or low grade bladder cancer (P=-0.002,0.003 and 0.007 P=0.014,0.002; and 0.033). D*, F and alpha value in the differential diagnosis of bladder cancer staging and histological grading was no statistically significant difference .ADC, D and DDC value in differential diagnosis of bladder cancer staging area under the ROC curve (the areas under the receiver operating characteristic curves, Az) were 0.751,0.747 and 0.726 in the differential diagnosis of bladder cancer: histological grading of Az were 0.733,0.796 and 0.702 respectively; ADC, D and DDC value in differential diagnosis of bladder cancer staging and grading 22 Az were no statistically significant difference. Conclusion: ADC, D and DDC can effectively identify muscle invasive bladder cancer is still non muscle invasive, but also can effectively identify bladder cancer is high level or low level; however, their Az diagnostic efficacy had no significant difference. DWI imaging three index model can be used to effectively identify the grade of bladder cancer staging and organization. The third part of the BOLD R2* imaging value in bladder cancer staging and grading of the objective preliminary study: the purpose of this study is to explore the Discuss the BOLD imaging R2* value in differential diagnosis of bladder cancer staging and grading. Materials and methods: This study by Tongji Hospital Affiliated to Tongji Medical College of Huazhong University of Science and Technology ethics committee approval, all the enrolled patients signed the informed consent.92 patients suspected bladder cancer underwent routine MRI scan and T2*Mapping sequence scanning. After the final screening of patients for 57 cases of.57 patients with a total of 90 bladder cancer lesions, all bladder cancer lesions were obtained in histological grade and tissue pathological staging results. Two diagnostic radiology scholars data R2* value measurement of bladder cancer in unaware of the pathological results under the condition of the two examiners. Between the measured value of consistency analysis by Bland Altman. Mann-Whitney U test to compare different bladder cancer staging and grading of R2* value difference. In addition, work with the subjects Curve (receiver operating characteristic, ROC) to determine the quantitative index of each invasive and non muscle invasive bladder cancer diagnostic efficacy in the diagnosis of bladder cancer and high level muscle. Results: the consistency between the two examiners better. High, low level of bladder cancer R2* = 20.97 + 8.91Hz. 16.43 + 5.74 Hz, two of them have statistically significant difference (P=-0.002); muscle invasive and non muscle invasive bladder cancer: R2* = 21.35 + 8.96Hz, 18.48 + 7.71Hz, two with statistically significant difference between the values of.R2* (P=-0.046) ROC curve area differential diagnosis of high and low grade of bladder cancer was 0.703 identification, myometrial invasion ROC curve area of bladder cancer and non muscle invasive bladder cancer was 0.636. conclusion: BOLD series R2* value can well identify classification and staging of bladder cancer, can for clinical treatment of bladder cancer clinicians The choice of the scheme provides favorable information.

【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：R445.2;R737.14

【相似文献】