Magnetic resonance imaging (MRI)-based intratumoral and peritumoral radiomics for prognosis prediction in glioma patients

Highlights•Combined intratumoral and peritumoral radiomics predict glioma prognosis.•Insights into peritumor microenvironment aid individualized surgical plans.•T2WI strongly correlates with glioma prognosis, more than T1-enhanced sequences.•T2WI provides richer content for future multimodal MRI research.AbstractObjectiveThe purpose of this study was to identify robust radiological features from intratumoral and peritumoral regions, evaluate MRI protocols and machine learning methods for overall survival stratification of glioma patients, and explore the relationship between radiological features and the tumor microenvironment.MethodsA retrospective analysis was conducted on 163 glioma patients, divided into a training set (n=113) and a testing set (n=50). For each patient, 2135 features were extracted from clinical MRI. Feature selection was performed using the Minimum Redundancy Maximum Relevance method and the Random Forest (RF) algorithm. Prognostic factors were assessed using the Cox proportional hazards model. Four machine learning models (RF, Logistic Regression, Support Vector Machine, and XGBoost) were trained on clinical and radiological features from tumor and peritumoral regions. Model evaluations on the testing set used Receiver Operating Characteristic curves.ResultsAmong the 163 patients, 96 had an overall survival (OS) of less than three years post-surgery, while 67 had an OS of more than three years. Univariate Cox regression in the validation set indicated that age (p=0.003) and tumor grade (p<0.001) were positively associated with the risk of death within three years post-surgery. The final predictive model incorporated 13 radiological and 7 clinical features. The RF model, combining intratumor and peritumor radiomics, achieved the best predictive performance (AUC = 0.91; ACC = 0.86), outperforming single-region models.ConclusionCombined intratumoral and peritumoral radiomics can improve survival prediction and has potential as a practical imaging biomarker to guide clinical decision-making.