Two-Stage Training Framework Using Multicontrast MRI Radiomics for IDH Mutation Status Prediction in Glioma

Abstract

Purpose To develop a radiomics framework for preoperative MRI-based prediction of isocitrate dehydrogenase (IDH) mutation status, a crucial glioma prognostic indicator. Materials and Methods Radiomics features (shape, first-order statistics, and texture) were extracted from the whole tumor or the combination of nonenhancing, necrosis, and edema regions. Segmentation masks were obtained via the federated tumor segmentation tool or the original data source. Boruta, a wrapper-based feature selection algorithm, identified relevant features. Addressing the imbalance between mutated and wild-type cases, multiple prediction models were trained on balanced data subsets using random forest or XGBoost and assembled to build the final classifier. The framework was evaluated using retrospective MRI scans from three public datasets (The Cancer Imaging Archive [TCIA, 227 patients], the University of California San Francisco Preoperative Diffuse Glioma MRI dataset [UCSF, 495 patients], and the Erasmus Glioma Database [EGD, 456 patients]) and internal datasets collected from the University of Texas Southwestern Medical Center (UTSW, 356 patients), New York University (NYU, 136 patients), and University of Wisconsin-Madison (UWM, 174 patients). TCIA and UTSW served as separate training sets, while the remaining data constituted the test set (1617 or 1488 testing cases, respectively). Results The best performing models trained on the TCIA dataset achieved area under the receiver operating characteristic curve (AUC) values of 0.89 for UTSW, 0.86 for NYU, 0.93 for UWM, 0.94 for UCSF, and 0.88 for EGD test sets. The best performing models trained on the UTSW dataset achieved slightly higher AUCs: 0.92 for TCIA, 0.88 for NYU, 0.96 for UWM, 0.93 for UCSF, and 0.90 for EGD. Conclusion This MRI radiomics-based framework shows promise for accurate preoperative prediction of IDH mutation status in patients with glioma. Keywords: Glioma, Isocitrate Dehydrogenase Mutation, IDH Mutation, Radiomics, MRI Supplemental material is available for this article. Published under a CC BY 4.0 license. See also commentary by Moassefi and Erickson in this issue.

Keywords: Glioma; IDH Mutation; Isocitrate Dehydrogenase Mutation; MRI; Radiomics.

Graphical abstract

Figure 1:

Flowchart of proposed MRI radiomics-based framework for predicting isocitrate dehydrogenase (IDH) mutation status in gliomas.

Figure 2:

Receiver operating characteristic (ROC) curves for the random forest and XGBoost (XGB) models trained using relevant features from the whole tumor; the nonenhancing tumor (NET), necrosis (NCR), and edema (ED) region of interest; and the combined features from both regions of interest. (A) ROC curves for the combined test sets from the University of Texas Southwestern Medical Center (UTSW), New York University (NYU), University of Wisconsin–Madison, and University of California San Francisco Preoperative Diffuse Glioma MRI dataset (UCSF) obtained by models trained on The Cancer Imaging Archive (TCIA) dataset. (B) ROC curves for the combined test sets from TCIA, UTSW, NYU, and UCSF obtained by models trained on the UTSW dataset. AUC = area under the receiver operating characteristic curve, RF = random forest.

Figure 3:

Summary of the most frequently selected image types, feature classes, and MRI sequences. (A) The most selected image types. (B) The most selected feature classes. (C) The percentage of MRI sequences identified as relevant features. GLCM = gray-level co-occurrence matrix, GLDM = gray-level dependence matrix, GLRLM = gray-level run-length matrix, GLSZM = gray-level size zone matrix, LoG = Laplacian of Gaussian, NGTDM = neighboring gray-tone difference matrix, WL-HH = wavelet filtering with high-pass filters, WL-HL = wavelet filtering with high-pass and low-pass filters, WL-LH = wavelet filtering with low-pass and high-pass filters, WL-LL = wavelet filtering with low-pass filters.

Figure 4:

Chart of the top 20 relevant features with their importance scores measured by Shapley additive explanations (SHAP) values. The feature names are labeled as region of interest 1 (ROI1) and ROI2, representing the whole tumor and the nonenhancing tumor (NET), necrosis (NCR), and edema (ED) region of interests, respectively. glcm = gray-level co-occurrence matrix, glszm = gray-level size zone matrix, log = Laplacian of Gaussian, ROI = region of interest, wavelet-LH = wavelet filtering with low-pass and high-pass filters, wavelet-LL = wavelet filtering with low-pass filters.