The Consistency Analysis And Combination Use Of Different Models Of MR Perfusion Weighted Imaging In The Classification Of Glioma

For now, perfusion weighted magnetic resonance imaging(PWI-MRI) is one of the most advanced technologies for the assessment and classification of glioma, which could be used to evaluate blood perfusion state and vascular permeability through analyzing the functional changes and hemodynamics of lesions. PWI-MRI mainly divided into three categories: Dynamic Susceptibility Contrast enhancement, Dynamic Contrast enhancement, Arterial spin labeling. The main imaging parameters of PWI-MRI include relative cerebral blood Volume(r CBV), relative cerebral blood flow(r CBF), volume transfer constant(Ktrans) and volume fraction of extra-vascular extracellular space(Ve). T*2 weighted dynamic magnetic sensitive contrast(dynamic susceptibility contrast DSC) MRI can provides hemodynamic information of intracranial tumors, which is the most widely used method in the classification of glioma pathology and has great value in the diagnosis and classification of glioma. At the same time, both the current domestic and foreign studies have shown that the relative cerebral blood volume(r CBV) can be used to evaluate and measure tumor angiogenesis of glioma. Some researchers believe that rCBV cannot be used as a reference parameters for III and IV gliomas grade diagnosis. Ktrans is also applied into the classification of gliomas, but it also exists uncertainty in the aspect of physics technology. The leakage of calculation could delay the DCE MRI, it will affect the accuracy of Ktrans and Ve values. Compared with DSC- MRI, DCE MRI can reflect the vascular permeability which can provide a more complete quantitative parameters in the evaluation of brain tumor angiogenesis.DSC perfusion imaging and DCE perfusion imaging have a variety of models: Compartment model, Patlak model, St. Lawrence-Lee model and Tofts model. There are few researches about this two technologies in the classification of gliomas, therefore, this paper just want to evaluate the combination use of DCE-MRI and DSC-MRI in the classification of glioma through different pharmacokinetic models, and to clear the application advantages of this two perfusion imaging technologies.Purpose: The combined application of different models of DSC-MRI and DCE-MRI in the classification of glioma, to study on the consistency analysis of results which calculated by the different models of DSC-MRI and DCE-MRI, to evaluate the accuracy and reliability of PWI-MRI in clinical application, to provide a foundation for improving the diagnosis of classification of glioma by using PWI-MRI.Methods:1. Patients and samples90 patients with glioma confirmed by histopathology were enrolled in this study, including 42 men and 48 women, and median age, 45 years(range, 12-72 years). All patients were first diagnosed and did not undergo any invasive or non-invasive treatment. According to the classification in WHO central nervous system(CNS), 32 patients with low-grade glioma and 58 patients with high-grade glioma were included.2. Equipment and parameterRoutine MRI scan, DCE and DSC scan were carried out by using SIEMENS Magnetom Verio 3.0T.The maximal values of Ktrans、Ve、r CBV and r CBF were measured in gliomas, respectively.3. Data analysesAll data was analyzed by SPSS19.0 software package. A P value of less than 0.05 was considered to be statistically significant. Receiver operating characteristic(ROC) curve was used to determine the cutoff of maximal values of Ktrans、Ve、rCBV and r CBF in distinguishing the patients with low-grade glioma from those with high-grade glioma, and the sensitivity and specificity were determined.Results:The mean values of Ktrans and Ve in HGG were both significent higher than that of LGG,(both P <0.05).The threshold value of Ktransand and Ve provided the best combination of sensitivity and specificity in differentiation between LGG and HGG. The mean values of r CBV and r CBF in HGG were both significent higher than that of LGG(both P <0.05).The threshold value of r CBV and rCBF provided the best combination of sensitivity and specificity in differentiation between LGG and HGG. The combination of r CBV with Ktrans can improve the sensibility and specificity in different models.Conclusion:rCBV, r CBF, Ktrans and Ve measurements correlated well with histopathologic grade. r CBV was the best predictor of glioma grade, and the combination of r CBV with Ktrans was the best set of metrics to predict glioma grade.