The Feasibility Study Of Predicting MGMT Methylation And IDH1 Gene Mutation In Glioma Based On MRI Imaging And Radiomics

Objective Based on the imaging features of glioma MRI-enhanced images,predict its MGMT and IDH1 gene status,in order to use the imaging features of gliomas to predict relevant immune indicators,and evaluate the prognosis and treatment effects of gliomas.Methods The pathological and imaging data of 99 patients with glioma confirmed by pathology in Ning Xia Medical University General Hospital were retrospectively collected.Use ITK-SNAP to delineate the region of interest(ROI)on the enhanced image,and perform imageomics analysis on it on the Artificial Intelligence Kit(GE Healthcare,Life Science,China).Use the LASSO regression model to select the key features of the MGMT and IDH1 genes to establish a model.The receiver operating characteristic(ROC)curve,area under the curve(AUC),sensitivity and specificity are used to evaluate the performance of the model in predicting the expression of related indicators.Results 1.The average age of IDH1 wild-type patients(51.45±14.02)is higher than the average age of IDH1 mutant patients(45.49±11.84);IDH1 wild-type patients have a higher degree of reinforcement than IDH1 mutant patients,and the differences between groups are statistically significant(P<0.05).There was a statistically significant difference in tumor level between MGMT methylated and non-methylated patients(P<0.05).2.According to the IDH gene status,select 7 key features on the enhanced image for modeling.The prediction model AUC value is 0.87,positive predictive value is 0.77,and negative predictive value is 0.77,specificity is 0.75,sensitivity is 0.78.According to the expression level of MGMT,three key features were selected on the MRI-enhanced image for modeling.The AUC value of the prediction model was 0.89,the positive predictive value was 0.91,and the negative predictive value was 0.82,the specificity was 0.53,the sensitivity was 0.98.Conclusion The radiomic features based on MRI-enhanced images can predict the expression of MGMT and IDH1 in gliomas;it can provide valuable information for formulating treatment plans in clinical diagnosis and treatment activities and evaluating the effects of patients after treatment.