Non-Invasive MGMT Status Prediction in GBM Cancer Using Magnetic Resonance Images Radiomics Features: Univariate and Multivariate Radiogenomics Analysis

Background This study aimed to predict methylation status of the O6 methylguanine-DNA methyltransferase (MGMT) gene promoter status by using magnetic resonance imaging radiomics features, as well as univariate and multivariate analysis. Methods Eighty-two patients who had an MGMT methylation status were included in this study. Tumors were manually segmented in the 4 regions of magnetic resonance images, 1) whole tumor, 2) active/enhanced region, 3) necrotic regions, and 4) edema regions. About 7000 radiomics features were extracted for each patient. Feature selection and classifier were used to predict MGMT status through different machine learning algorithms. The area under the curve (AUC) of the receiver operating characteristic curve was used for model evaluations. Results Regarding univariate analysis, the Inverse Variance feature From Gray Level Co-occurrence Matrix in whole tumor segment with 4.5 mm Sigma of Laplacian of Gaussian filter with AUC of 0.71 (P value = 0.002) was found to be the best predictor. For multivariate analysis, the Decision Tree classifier with Select from Model feature selector and LOG (Laplacian of Gaussian) filter in edema region had the highest performance (AUC, 0.78), followed by Ada-Boost classifier with Select from Model feature selector and LOG filter in edema region (AUC, 0.74). Conclusions This study showed that radiomics using machine learning algorithms is a feasible noninvasive approach to predict MGMT methylation status in patients with glioblastoma multiforme cancer.