Deep Learning Radiomics to Predict PTEN Mutation Status From Magnetic Resonance Imaging in Patients With Glioma

Objectives: Phosphatase and tensin homolog (PTEN) mutation is an indicator of poor prognosis of low-grade and high-grade glioma. This study built a reliable model from multi-parametric magnetic resonance imaging (MRI) for predicting PTEN mutation status in patients with glioma. Methods: In this study, a total of 244 patients with glioma was retrospectively collected from our center (n=77) and The Cancer Imaging Archive (n=167). All patients were randomly divided into training (n=170) and validation (n=74) sets. Three models were built from preoperative MRI for predicting PTEN status, including a radiomics model, a convolutional neural network (CNN) model, and an integrated model based on both radiomics and CNN features. The performances of each model were evaluated by accuracy and area under the receiver operating characteristic curve (AUC). Results: The CNN model achieved an AUC of 0.84 and an accuracy of 0.81, which performed better than the radiomics model with an AUC of 0.83 and an accuracy of 0.66. Combing radiomics with CNN will further benefit the prediction performance (accuracy=0.86, AUC=0.91). Conclusions: Combination of both CNN and radiomics features achieved significantly higher performance than radiomics or CNN alone in predicting PTEN mutation status in patients with glioma.