MAGORINO: Magnitude-only fat fraction and R2* estimation with Rician noise modelling

Purpose: Magnitude-based fitting of chemical shift-encoded data enables proton density fat fraction (PDFF) and R2* estimation where complex-based methods fail or when phase data is inaccessible or unreliable, such as in multi-centre studies. However, traditional magnitude-based fitting algorithms suffer from Rician noise-related bias and fat-water swaps. To address these issues, we propose an algorithm for Magnitude-Only PDFF and R2* estimation with Rician Noise modelling (MAGORINO). Methods: Simulations of multi-echo gradient echo signal intensities are used to investigate the performance and behavior of MAGORINO over the space of clinically plausible PDFF, R2* and SNR values. Fitting performance is assessed in terms of parameter bias, precision and fitting error. To gain deeper insights into algorithm behavior, the paths on the likelihood functions are visualized and statistics describing correct optimization are generated. MAGORINO is compared against Gaussian noise-based magnitude fitting and complex fitting. Results: Simulations show that MAGORINO reduces bias in both PDFF and R2* measurements compared to Gaussian fitting, through two main mechanisms: (i) a greater chance of selecting the true (non-swapped) optimum, and (ii) a shift in the position of the optima such that the estimates are closer to ground truth solutions, as a result of the correct noise model. Conclusion: MAGORINO reduces fat-water swaps and Rician noise-related bias in PDFF and R2* estimation, thus addressing key limitations of traditional Gaussian noise-based magnitude-only fitting.