Development and Validation of an Efficient MRI Radiomics Signature for Improving the Predictive Performance of 1p/19q Co-Deletion in Lower-Grade Gliomas

Abstract

The prognosis and treatment plans for patients diagnosed with low-grade gliomas (LGGs) may significantly be improved if there is evidence of chromosome 1p/19q co-deletion mutation. Many studies proved that the codeletion status of 1p/19q enhances the sensitivity of the tumor to different types of therapeutics. However, the current clinical gold standard of detecting this chromosomal mutation remains invasive and poses implicit risks to patients. Radiomics features derived from medical images have been used as a new approach for non-invasive diagnosis and clinical decisions. This study proposed an eXtreme Gradient Boosting (XGBoost)-based model to predict the 1p/19q codeletion status in a binary classification task. We trained our model on the public database extracted from The Cancer Imaging Archive (TCIA), including 159 LGG patients with 1p/19q co-deletion mutation status. The XGBoost was the baseline algorithm, and we combined the SHapley Additive exPlanations (SHAP) analysis to select the seven most optimal radiomics features to build the final predictive model. Our final model achieved an accuracy of 87% and 82.8% on the training set and external test set, respectively. With seven wavelet radiomics features, our XGBoost-based model can identify the 1p/19q codeletion status in LGG-diagnosed patients for better management and address the drawbacks of invasive gold-standard tests in clinical practice.

A Radiomics-Based Machine Learning Model for Prediction of Tumor Mutational Burden in Lower-Grade Gliomas

“…For instance, Le et al [ 25 ] used the Spearman correlation test and F-score analysis to find significant features for glioblastoma identification. Kha et al [ 26 ] employed SHAP analysis, a machine learning model, to find a radiomics signature for 1p/19q codeletion status prediction. A novel method for feature selection was conducted in our study where an incorporated model of the genetic algorithm (GA) and machine learning algorithms was developed to identify the radiomics signature.…”

Utilization of model-agnostic explainable artificial intelligence frameworks in oncology: a narrative review

“…A challenge XAI helps address is identifying only the most important features predictive of the outcome, which improves final model feature selection, thus eliminating the irrelevant ones, leading to a better model that is more robust and less prone to overfitting. Kha et al have reported on identification of a radiomic signature predictive of 1p/19q co-deletion ( 34 ). During model development, the SHAP framework was used to identify the most important features to include in their model, leading to improved model performance when less important features, based on average SHAP values, were removed, with an AUC of 0.710 before feature selection and 0.753 after.…”

A Survey of Radiomics in Precision Diagnosis and Treatment of Adult Gliomas

“…1p/19q co-deletion refers to the combined deletion of the short arm of chromosome 1 and the long arm of chromosome 19, which can occur in various glioma subtypes, with oligodendroglioma being the most common, and GBM having a very low incidence. Many studies have demonstrated that 1p/19q co-deletion status enhances tumor sensitivity to different types of treatments [61]. Therefore, gliomas with 1p/19q co-deletion are sensitive to chemotherapy and have a significantly improved prognosis.…”