The Study Of Dynamic Contrast Enhanced Magnetic Resonance Imaging Histogram In The Differential Diagnosis Of High Grade Glioma And Brain Metastatic Tumor Of Lung Cancer

Objective: To explore the diagnostic value of whole field quantitative parameter histogram analysis in dynamic contrast enhanced magnetic resonance(DCE-MRI)study in the differential diagnosis of high-grade gliomas and metastatic tumors of lung cancer.Methods:A total of 23 patients with lung cancer brain metastases and high-grade gliomas examined by DCE-MRI were prospectively collected,including 14 cases of brain metastases from lung cancer and 9 cases of high-grade gliomas confirmed by operation and pathology.We use Omini Kinetics software and the application of Toft-extended postprocessing hemodynamics model to calculate the quantitative parameters of each case,including volume transfer constant(Ktrans)and reflux rate constant(Kep),extravascular extracellular volume ratio(Ve)and plasma volume ratio(Vp).The quantitative parameters of whole-lesion histogram analysis method at every level contains the tumor along the edge of tumor manually depicting ROI,the ROIs merged at all levels as a VOI,the software will automatically calculate the average of quantitative parameters(Ktrans value,kep value,ve value,VP value)and the histogram parameters of them,including skewness,kurtosis,energy,entropy,percentile(Quantile5,Quantile10,Quantile25,Quantile50,Quantile75,Quantile90,Quantile95).SPSS version 22 is used for data analysis: When the data accord with normal distribution and homogeneity of variance,two independent samples t test will be used to represent the result with "mean ± standard deviation".When the data does not conform to normal distribution or homogeneity of variance,The Mann-Whitney U test will be used and the result is expressed by "median four ± interval".The diagnostic efficiency of statistical difference parameters was compared by ROC curve.Results: 1.The mean value of Ve in the whole lesion of the brain metastatic tumor of lung cancer was higher than that of the high grade glioma,and the difference was statistically significant(P=0.044).The mean values of Ktrans,Kep and Vp of the brain metastatic tumors of lung cancer were slightly higher than those of high grade glioma,but the difference was not statistically significant(P > 0.05).2.The energy,entropy,Quantile 5 and Quantile 10 of the Ktrans histogram parameters were significantly different between the two groups(p=0.001,0.001,0.003,0.003).There were significant differences in energy,entropy,Quantile 5,Quantile 10 and Quantile 25 between the two groups in Kep histogram parameters(p=0.012,0.023,0.008,0.006,0.026).There were significant differences in the Ve histogram parameters,including skewness,consistency,energy,entropy,Quantile 5,Quantile 10 and Quantile 25 between the two groups(p=0.021,0.029,0.006,0.008,0.003,0.003,0.023,0.062).There were significant differences in the skewness,kurtosis,energy and entropy between the two groups in the Vp histogram parameters(p=0.006,0.014,0.042,0.042).3.Ktrans-Quantile 5,Ktrans-Quantile 10,Kep-Quantile 5,Kep-Quantile 10,KepQuantile 25,Ve-Quantile 5,Ve-Quantile 10,Ve-Quantile 25,Ktrans-entropy,Kep-entropy,Ve-entropy,Ve-uniformity had differential diagnosis efficiency.The Ktrans-entropy had the highest diagnosis efficiency.Conclusion:The whole lesion quantitative parameter histogram of DCE-MRI can be used as a clinical application to identify high grade glioma and brain metastatic tumor of the brain lung cancer.