
A study on 114 glioma patients examines the utility of multiparametric MRI radiomics in preoperatively predicting isocitrate dehydrogenase (IDH) genotype. Patient data is split into training (n=79) and validation cohorts (n=35). After feature extraction, the radiomics model, incorporating eight selected features, demonstrates superior performance (AUC 0.974 and 0.872) compared to conventional models. Integrating clinical data, the combined model further enhances predictive accuracy (AUC 0.963 and 0.892). The findings suggest that multiparametric MRI radiomics models can aid in preoperative IDH genotype prediction for gliomas.
Like
Save
Share