Improved diagnostic performance of APTw MRI to multiparametric non-contrast-enhanced MRIs in patients with post-treatment high-grade gliomas

We explored non-contrast-enhanced MRI performances of APTw, DWI, SWI, and pCASL at 3 Tesla in glioma patients post-treatment. APTw, ADC, QSM, and CBF histogram parameters from volumetric ROIs were recorded. Multivariable logistics regression with principal component analysis (PCA) was built for differentiating treatment effect from tumor recurrence. Results showed that the regression model trained on the combination of APTw, CBF, and QSM achieved the highest classification performance, with an AUC of 0.90.