| Part I A comparative study of3-dimensional pseudocontinuousarterial spin labeling and dynamic susceptibility contrast-enhancedperfusion weighted-imaging in neuroepithelial tumors gradingObjective: To retrospectively analyses the diagnostic accuracy of3-dimensionalpseudocontinous arterial spin labeling(3D-PCASL) and dynamic susceptibilitycontrast(DSC)-enhanced perfusion in the prediction of neuroepithelial tumors grading,and to appraise normalized lesion/normal tissue cerebral blood flow (nCBF) andnormalized lesion/normal tissue relative cerebral blood flow(nrCBF) ofneuroepithelial tumors with histological findings as a reference standard.Material and Methods:62patients with pathologically proved neuroepithelialtumors underwent conventional magnetic resonance sequences,3-dimensionalpseudocontinuous arterial spin labeling(3D-PCASL) and dynamic susceptibilitycontrast(DSC)-enhanced perfusion MR imaging. Representative maximal nCBForiginated from3D-PCASL and nrCBF from DSC regions of interest were chosenfrom each lesion. These parameters were used for statistical evaluation comparingwith histopathological tumor grade.Results: The mean nCBF ratio (1.796±1.289)of low grade tumor was significantlylower than that (4.329±2.046) of high grade group（P＜0.05）. The mean nrCBFratio(1.778±1.209) of low trade group was also significantly lower than that(5.927±3.448) of high grade group（P＜0.05）. There was a close positive correlationbetween nCBF and nrCBF in the tumor region of interest (linear regression coefficient,R=0.790; P＜.0001).Conclusion：Accounting for its intrinsic advantages,3D-PCASL is a alternative or prefered method of DSC for assessment of microvascular perfusion and allowsdistinction between high-and low-grade neuroepithelial tumors, especially for thosepatients required following-up or possessed contraindication of contrast. Part II Application of multiple b values diffusion-weighted imagingfor neuroepithelial tumors gradingObjective: With DSC perfusion-weighted imaging as reference standards, toretrospectively appraise the diagnostic efficacy of perfusion-related parameters ofmultiple b values diffusion weighted image (DWI) in the prediction of neuroepithelialtumors grading. With MinADC from standard DWI as reference standards, toretrospectively appraise the diagnostic efficacy of diffusion-related parameters ofmultiple b values diffusion weighted image (DWI) in the prediction of neuroepithelialtumors grading.Material and Methods:62patients with pathologically proved neuroepithelialtumors underwent conventional magnetic resonance sequences, multiple b values(0～4000s/mm2,totally12b values) diffusion weighted image (DWI), dynamicsusceptibility contrast-enhanced perfusion MR imaging. With tumor hot-spot area inrCBF map as reference standards, the multiple perfusion-related parameters of tumorswere measured postoperatively by using parameter maps progressing frommono-exponential and biexponential model of multiple b values DWI. With MinADCarea from standard DWI as a reference standard, the multiple diffusion-relatedparameters of tumors were measured postoperatively by using parameter mapsprogressing from mono-exponential, biexponential, and stretched model of multipleb values DWI.These parameters were used for statistical evaluation comparing withhistopathological tumor grade.Results：The mean fast ADC-mono （3.794±2.186）×10-3mm2/s of low grade tumorwas slightly lower than that （4.030±1.597）×10-3mm2/s of high grade group（P=0.612）. The mean fraction of fast ADC-mono （0.398±0.217）of low gradetumor was slightly lower than that（0.419±0.172）of high grade group（P=0.356）.Themean fast ADC-bi （10.578±14.261）×10-3mm2/s of low trade group was slightlylower than that （13.141±12.895）×10-3mm2/s of high grade group（P=0.462）. Themean fraction of fast ADC-bi （0.558±0.173）of low grade tumor was slightly lowerthan that （0.679±0.157）of high grade group（P=0.172）. The group differences werenot significant for these perfusion-related parameters of mono-exponential andbiexponential model of multiple b values DWI. The mean MinADC （0.939±0.436）×10-3mm2/s from standard DWI of low grade tumor was higher than that（0.560±0.191）×10-3mm2/s of high grade group（P=0.005）. The mean slowADC-mono （0.777±0.315）×10-3mm2/s of low grade tumor was higher than that（0.488±0.163）×10-3mm2/s of high grade group（P=0.004）. The mean slow ADC-bi（0.675±0.726）×10-3mm2/s of low grade group was higher than that（0.319±0.276）×10-3mm2/s of high grade group（P=0.091）. The mean slow DDC（1.200±0.670）×10-3mm2/s of low grade group was higher than that （0.617±0.235）×10-3mm2/s of high grade group（P=0.006）. There isnot difference between low tradegroup（0.878±0.085） and high grade group(0.834±0.107)（P=0.191）for mean alphavalues.Conclusion：There aren’t equal efficacy between those perfusion-related parametersof mono-exponential and biexponential model of multiple b values DWI and DSC. Itshould be elucidated in the futue research that those perfusion-related parameters ofmono-exponential and biexponential model of multiple b values DWI can assessmicrovascular perfusion of neuroepithelial tumors.The diffusion-related parameters ofof mono-exponential, biexponential and stretched exponential model of multiple bvalues DWI can present truer diffusion of neuroepithelial tumors than standard DWIdo. The difference between low-grade tumor (LGT) and high-grade tumor (HGT) wasmost significant by slow ADC-mono based on diffusion-related parameters ofmultiple b values DWI. Part III Application of Combining different perfusion-weightedimaging techniques and multiple b values diffusion-weighted imagingfor neuroepithelial tumors gradingObjective: Combining DSC and multiple b values DWI or3D-PCASL and multiple bvalues DWI to compare the diagnostic efficacy of bi-modalities to individualmodality in the prediction of neuroepithelial tumors grading, respectively.Material and Methods:62patients with pathologically proved neuroepithelialtumors underwent conventional magnetic resonance sequences,3D-PCASL, multipleb values(0～4000s/mm2,totally12b values) diffusion weighted image (DWI) andDSC. Withdraw the diffusion-related parameters of multiple b values DWI andparameters of DSC or3D-PCASL, using ROC and Pearson’s relationship analysis tocalculate the threshold diagnotic value for differentiating high-from slow-gradeneuroepithelial tumors, and to compare efficacy of combination of two modalities toindividual modality for grading two group tumors.Results：There was a close positive correlation between nCBF and nrCBF in the ROIof tumors (linear regression coefficient, R=0.790; P＜.0001). And There were aclose negative correlation between nCBF/nrCBF and slow ADC-mono in the ROI oftumors (linear regression coefficient, R=-0.519; P＜.0001or R=-0.537; P＜0.001).The cutoff nCBF ratio of2.380, nrCBF ratio of3.134and slow ADC-mono of0.578×10-3mm2/s for the differentiation between high and low grade neuroepithelial tumorsprovided the best combination of sensitivity (93.3%,96.7%,95.1%), specificity(81.3%,87.5%,64.3%) and accuracy (85.4%,91.9%,82.3%) respectively. Combiningthe cutoff value of slow ADC-mono with nCBF or nrCBF could improve the gradingsensitivity to96.7%and97.3%, specificity to94.4%and92.3%, accuracy to90.3%and93.7%, respectively.Conclusion：The combination of multiple b values DWI and3D-PCASL or DSC-PWIcan markedly improve the accuracy for neuroepithelial tumor grading, andparticularly the combination of contrast-free multiple b values DWI and3D-PCASLsequences can be a preferred and promising methods of choice. |
PDF Full Text Request