不同灌注成像技术与多b值扩散加权成像在脑神经上皮肿瘤分级评估中的研究

发布时间：2018-06-19 16:36

本文选题：动态磁敏感对比增强 + 三维准连续动脉自旋标记　；参考：《中国人民解放军医学院》2014年博士论文

【摘要】：第一部分动态磁敏感对比增强与三维准连续动脉自旋标记灌注成像在脑神经上皮肿瘤分级评估中的对照研究目的：通过比较高级别肿瘤组与低级别肿瘤组肿瘤最大血流量与正常小脑白质血流量的比值（nCBF）以及最大相对血流量与正常小脑白质血流量的比值（nrCBF）组间差异，回顾性分析动态磁敏感对比（DSC）增强灌注成像和三维准连续动脉自旋标记成像（3D-PCASL）在脑神经上皮肿瘤分级中的诊断准确性。材料与方法：62个经病理证实的神经上皮肿瘤病人术前行常规磁共振平扫、3D-PCASL、团注造影剂的同时行DSC灌注成像，然后行常规增强扫描。运用后处理图像，根据3D-PCASL-CBF图及DSC-rCBF图，术后测量肿瘤区最大CBF值及正常小脑白质区CBF值并求比值（nCBF），同样的方法测量nrCBF值，按照病理结果分高、低级别组进行相应参数的统计学分析和Pearson相关性分析。结果：低级别肿瘤组平均nCBF值（1.796±1.289）明显低于高级别组（4.329±2.046），差异具有统计学意义（P＜0.05）。低级别肿瘤组平均nrCBF值（1.778±1.209）明显低于高级别组（5.927±3.448），差异同样具有统计学意义（P＜0.05）。肿瘤nCBF与nrCBF值存在明显正相关关系，相关系数为0.790（P＜0.001）。结论：由于诸多内在的优势，3D-PCASL完全可以用于评价肿瘤微血管灌注情况，明确区分高级别及低级别神经上皮肿瘤，具有与DSC相当的诊断效能，在需要反复随访和不适宜用造影剂的病例，可作为DSC的优先选择方法。第二部分多b值扩散加权成像在脑神经上皮肿瘤分级评估中的应用目的：分别以标准DWI和动态磁敏感对比（DSC）增强灌注成像为对照，评价多b值扩散加权成像（DWI）扩散参数和灌注相关参数在神经上皮肿瘤分级中的应用价值。材料与方法：62个经病理证实的神经上皮肿瘤病人术前行常规磁共振、多b值DWI以及DSC检查、常规增强扫描。分别以肿瘤内rCBF热点区为参照，测量肿瘤内多b值DWI单指数模型、双指数模型与灌注相关参数fast ADC-mono、fraction of fastADC-mon、fast ADC-bi、fraction of fast ADC-bi值，以肿瘤内标准DWI中最低ADC值为参照，测量肿瘤内多b值DWI单指数模型、双指数模型与扩散相关参数slowADC-mono、slow ADC-bi值以及拉伸指数模型DDC和α值，按照病理结果分高低级别肿瘤组分别与相关参数进行组间统计学差异分析。结果：以rCBF值为参照，低级别组平均fast ADC-mono值（3.794±2.186）×10-3mm2/s稍低于高级别组（4.030±1.597）×10-3mm2/s，差异无明显统计学意义（P=0.612）。低级别组平均fraction of fast ADC-mono值（0.398±0.217）稍低于高级别组（0.419±0.172），差异无明显统计学意义（P=0.356）。低级别组平均fast ADC-bi值（10.578±14.261）×10-3mm2/s与高级别组（13.141±12.895）×10-3mm2/s无明显差异（P=0.462）。低级别组平均fraction of fast ADC-bi值（0.558±0.173）稍低于高级别组（0.679±0.157），差异无明显统计学意义（P=0.172）。标准DWI中低级别肿瘤组最小ADC值（0.939±0.436）×10-3mm2/s明显高于高级别组（0.560±0.191）×10-3mm2/s，差异有明显统计学意义（P=0.005）。以标准DWI中最小ADC值为参照，低级别组平均slow ADC-mono值（0.777±0.315）×10-3mm2/s明显高于高级别组（0.488±0.163）×10-3mm2/s，差异有明显统计学意义（P=0.004）。低级别组平均slow ADC-bi值（0.675±0.726）×10-3mm2/s高于高级别组（0.319±0.276）×10-3mm2/s，但无明显差异（P=0.091）。低级别组平均DDC值（1.200±0.670）×10-3mm2/s高于高级别组（0.617±0.235）×10-3mm2/s，差异具有明显统计学意义（P=0.006）。低级别组平均α值（0.878±0.085）与高级别组0.834±0.107差异无明显统计学意义（P=0.191）。结论：在神经上皮肿瘤分级中，多b值DWI单、双指数模型与灌注相关参数尚不具有分级价值，能否体现肿瘤的灌注情况还有待进一步实验证实。多b值DWI单、双指数模型与扩散相关参数及拉伸指数模型参数比标准ADC更能够体现肿瘤的真实扩散情况，其中以slow ADC-mono差异最明显。第三部分联合不同灌注成像技术与多b值扩散加权成像在脑神经上皮肿瘤分级评估中的应用目的：分别联合动态磁敏感对比增强（DSC）灌注成像与多b值DWI和三维准连续动脉自旋标记（3D-PCASL）灌注成像与多b值DWI成像方法，分析比较联合灌注与多b值DWI双模态技术与任何单一模态技术在神经上皮肿瘤中的分级效能。材料与方法：62个经病理证实的脑神经上皮肿瘤病人术前行常规磁共振、多b值DWI（0～4000s/mm2，一共12个b值）、3D-PCASL及DSC检查。提取前述有关灌注参数与扩散参数，按照高低级别肿瘤组分组分别进行ROC曲线分析及相关性分析，确定各自诊断阈值，分析评估各自及联合诊断效能。结果：如前所述，肿瘤nCBF与nrCBF值存在明显正相关关系，相关系数为0.790（P＜0.001），以nCBF值2.380和nrCBF值3.134作为阈值，，可分别达到93.3%、96.7%的诊断敏感性，81.3%、87.5%的特异性和85.4%、91.9%的准确性。肿瘤nCBF或nrCBF与slow ADC-mono值存在明显负相关关系，相关系数分别为R=-0.519（P＜0.001）和R=-0.537（P＜0.001），以slow ADC-mono值0.578×10-3mm2/s作为诊断阈值，能够达到95.1%的诊断敏感性，64.3%的特异性和82.3%的准确性。分别联合肿瘤nCBF或nrCBF与slow ADC-mono值，能够达到96.7%和97.3%诊断敏感性、94.4%和92.3%特异性、90.3%和93.7%的准确性。结论：分别联合两种灌注成像技术与多b值DWI可以明显提高脑神经上皮肿瘤的分级诊断效能，尤其是三维准连续动脉自旋标记灌注成像联合多b值DWI，由于不用造影剂而具有更加广泛的临床应用前景。
[Abstract]:Comparative study of dynamic magnetic susceptibility contrast enhancement and three - dimensional quasi - continuous artery spin - labeled perfusion imaging in the assessment of neuroepithelial tumors

Objective : To investigate the diagnostic accuracy of dynamic magnetic sensitivity contrast ( DSC ) - enhanced perfusion imaging and three - dimensional quasi - continuous arterial spin - labeled imaging ( 3D - PCASL ) in the classification of brain neuroepithelial tumors by comparing the ratio of maximum blood flow ( nCBF ) and the ratio of maximum relative blood flow ( nCBF ) and normal cerebellar white blood flow ( nCBF ) between high - grade tumor group and low - grade tumor group .

Materials and Methods : Sixty - two patients with neuroepithelial tumor proved by pathology underwent conventional magnetic resonance plain scan , 3D - PCASL , bolus contrast agent simultaneous scanning calorimetry ( DSC ) perfusion imaging , followed by routine enhanced scanning . Using post - processing image , the maximal CBF value and CBF value of normal cerebellum were measured and the ratio ( nCBF ) was measured in the same way . The statistical analysis of the parameters and Pearson correlation analysis were performed according to the pathological results .

Results : The mean nCBF value of low grade tumor group ( 1.796 卤 1.289 ) was significantly lower than that in the advanced group ( 4.329 卤 2.46 ) . The difference was statistically significant ( P < 0.05 ) .

Application of the second part of multi - b - valued diffusion - weighted imaging in the assessment of neuroepithelial tumors

Objective : To evaluate the value of diffusion - weighted imaging ( DWI ) diffusion parameters and perfusion - related parameters in the classification of neuroepithelial tumors by contrast with standard DWI and dynamic magnetic sensitive contrast ( DSC ) .

Materials and Methods : Sixty - two patients with neuroepithelial tumor proved by pathology were subjected to routine magnetic resonance , multi - b - value DWI , DSC and conventional enhancement scanning . The tumor was divided into multiple - index DWI single - index model , dual - exponential model and perfusion - related parameters fast ADC - bi , fraction of fast ADC - bi values .

Results : The average fast ADC - bi values of low - grade group were significantly higher than that in the advanced group ( P = 0 . 05 ) . The mean slow ADC - bi values ( 0 .

Conclusion : In the classification of neuroepithelial tumors , the multi - value DWI alone , biexponential model and perfusion - related parameters are not of the hierarchical value , and whether the perfusion of tumor can be demonstrated . The multi - b - value DWI single , double exponential model and diffusion - related parameters and tensile index model parameters can reflect the true diffusion of tumors more clearly than the standard ADC , with slow ADC - mono difference being the most obvious .

Application of the third part combined with different perfusion imaging techniques and multi - b - value diffusion weighted imaging in the assessment of brain neuroepithelial tumors

Objective : To compare the imaging of dynamic magnetic susceptibility contrast enhancement ( DSC ) perfusion imaging with multi - b - value DWI and 3D - PCASL perfusion imaging and multi - b - value DWI imaging .

Materials and Methods : Sixty - two patients with neuroepithelial tumors confirmed by pathology underwent conventional magnetic resonance , multi - b - value DWI ( 0 - 4000s / mm2 , total 12 b - values ) , 3D - PCASL and DSC . The above - mentioned perfusion parameters and diffusion parameters were extracted , ROC curve analysis and correlation analysis were carried out according to the tumor component groups at high and low levels , and the respective diagnostic thresholds were determined , and their respective diagnostic efficacy were analyzed and assessed .

Results : As mentioned earlier , there was a positive correlation between the nCBF values of tumor nCBF and the value of nCBF , the correlation coefficient was 0.790 ( P < 0.001 ) , and the accuracy of 95.3 % , 97.7 % , 81.3 % , 87.5 % specificity and 85.4 % , 91.9 % accuracy respectively . The correlation coefficients were as follows : R = - 0.519 ( P & lt ; 0.001 ) and R = - 0.537 ( P & lt ; 0.001 ) .

Conclusion : Combination of two perfusion imaging techniques and multi - b - value DWI can significantly improve the classification and diagnostic efficacy of brain neuroepithelial tumors , especially in three - dimensional quasi - continuous artery spin - labeled perfusion imaging combined with multi - b - value DWI , which has a wider clinical application prospect due to the use of contrast agents .
【学位授予单位】：中国人民解放军医学院
【学位级别】：博士
【学位授予年份】：2014
【分类号】：R739.41;R445.2

【相似文献】