We implement generative adversarial network (GAN) models to predict fully quantitative parameters from a complex-valued multiecho MRI sequence using only data from the magnitude-only two-point Dixon acquisition in the UK Biobank abdominal protocol. The training data consists of in- and opposed-phase channels from the Dixon sequence as inputs and the proton density fat fraction (PDFF) and R2* parameter maps estimated from the IDEAL acquisition as outputs. We compare conditional and cycle GANs, where the conditional GAN models outperformed the cycleGAN models in SSIM, PSNR and MSE. PDFF predictions were better than R2* predictions for all models.
This abstract and the presentation materials are available to members only; a login is required.