Prediction of IDH status of Glioma using Diffusion Tensor Imaging and Clinical features

Isocitrate dehydrogenase (IDH) is critical to prognosis of glioma. While, reliable techniques for preoperative assessment of IDH status remain scarce. In this study, we investigated mean diffusivity (MD) and anisotropy fraction (FA) using Diffusion Tensor Imaging (DTI) combined with the clinical features to predict IDH status. Our results found significant differences in FAmean/FAnawm, MDmin, NLR, and age between IDH mutant and IDH wild groups. The model incorporating FAmean/FAnawm, MDmin, NLR, and age predicted IDH status with area under ROC curve of 0.85, 95% CI: 74.3%~95.7%. Our findings suggested that DTI combined with clinical features can non-invasively prediction of IDH status.