Quantification of NOE effect and its application on brain tumor detection

Chemical exchange saturation transfer (CEST) quantification is mostly based on pixel-wise fitting of Z-spectra, which is time-consuming and noise-sensitive. Herein, we propose an approach to quantify CEST based on U-Net. The proposed method can simultaneously quantify the concentration and the exchange rate of nuclear Overhauser enhancement (NOE), together with the B0 map. The results of a simulation sample and a rat C6 glioma model suggest that the quantification of NOE effect with U-Net is accurate, precise and fast.