Machine Learning-Based Preoperative Predicting <i>TERT</i> Promoter Mutation and <i>EGFR</i> Gene Amplification Phenotype in <i>IDH</i> Wild-Type Glioblastoma Using Advanced MR Habitat Imaging.
Authors
Affiliations (5)
Affiliations (5)
- From the Department of Radiology (Y.S., W.G., Y.P., P.H., D.S.), The First Affiliated Hospital, Fujian Medical University, Fuzhou, Fujian, P.R. China.
- Department of Radiology (Y.S., W.G.), National Regional Medical Center, Binhai Campus of The First Affiliated Hospital, Fujian Medical University, Fuzhou, Fujian, P.R. China.
- MR Research Collaboration Team (Y.S.), Siemens Healthineers Ltd, Shanghai, China.
- From the Department of Radiology (Y.S., W.G., Y.P., P.H., D.S.), The First Affiliated Hospital, Fujian Medical University, Fuzhou, Fujian, P.R. China [email protected].
- Key Laboratory of Radiation Biology of Fujian Higher Education Institutions (D.S.), The First Affiliated Hospital, Fujian Medical University, Fuzhou, Fujian, P.R. China.
Abstract
The telomerase reverse transcriptase (<i>TERT</i>) gene promoter mutation is a crucial factor for identifying an isocitrate dehydrogenase (<i>IDH</i>) wild-type glioblastoma with poor prognosis, and the epidermal growth factor receptor (<i>EGFR</i>) amplification may be a potential prognostic factor. The purpose of this study was to investigate the value of the tumor habitats imaging model on advanced MRI in predicting <i>TERT</i> promoter mutation and <i>EGFR</i> gene amplification phenotype of <i>IDH</i> wild-type glioblastoma. One hundred seventy-nine patients with pretreatment conventional MRI, DWI, and DSC-PWI were included. The data were divided into the training set (<i>n</i>=112), test set (<i>n</i>=29), and time-independent validation set (<i>n</i>=38). Based on the ADC and CBV map, the solid tumor area was split into several habitat subregions using the k-means clustering algorithm (hypovascular hypercellular area, hypervascular area, and hypovascular hypocellular area). In the training set, <i>TERT</i> promoter mutation and <i>EGFR</i> gene amplification phenotype prediction models were constructed using the random forest method. The reliability of prediction models was validated in the test and the time-independent validation sets. Receiver operating characteristic (ROC) curve analysis, calibration curve, and decision curve analysis (DCA) were used. The area under the curve (AUC) of the training, test, and validation sets of the <i>TERT</i> promoter prediction model was 0.877, 0.783, and 0.796, respectively. The accuracy of the <i>TERT</i> promoter prediction model was 82.1%, 75.9%, and 76.3%, respectively. The AUCs of the 3 sets for the <i>EGFR</i> gene amplification status prediction model were 0.877, 0.784, and 0.878, respectively. The accuracy of the <i>EGFR</i> gene amplification status prediction model was 79.5%, 75.9%, and 89.5%, respectively. Moreover, the prediction probability of these models was in good agreement with the actual result. The tumor habitat imaging model based on advanced MRI was useful for accurately predicting <i>TERT</i> promoter mutation and <i>EGFR</i> amplification status in <i>IDH</i> wild-type glioblastoma.