Search papers, labs, and topics across Lattice.
This study evaluates the feasibility of estimating magnetic resonance imaging-signal fat fraction (MRI-SFF) from standard T2-weighted MRI images using image-to-image translation techniques. A lightweight 4-level U-Net model was compared against a state-of-the-art Denoising Diffusion Probabilistic Model (DDPM) using a large dataset of paired images, revealing that the U-Net not only outperformed the DDPM in terms of correlation and error but also significantly reduced inference time. The findings suggest that the U-Net can facilitate real-time clinical applications for MRI-SFF estimation, making it a practical alternative to more complex models.
A lightweight U-Net achieves superior performance to diffusion models in estimating MRI-SFF while slashing inference time by over 200 times.
Magnetic resonance imaging-signal fat fraction (MRI-SFF) quantifies tissue fat and serves as an established biomarker for metabolic and musculoskeletal disorders. The acquisition requires, however, specialized MRI sequences, which are not available routinely. We investigate whether SFF can be estimated from widely available T2-weighted (T2w) MRI via image-to-image translation (I2I). We further compare a lightweight 4-level U-Net to a state-of-the-art Denoising Diffusion Probabilistic Model (DDPM) using a dataset of 230 048 paired 2D images (183 517 train, 23 621 val, 22 910 test) from the German National Cohort (NAKO). Both models clearly outperform the identity baseline (Pearson correlation r = 0.769, mean absolute error MAE = 0.070 +/- 0.054), which confirms that the models learn a non-trivial cross-modal mapping. Interestingly, the lightweight U-Net outperforms the DDPM in both correlation (r = 0.975 vs. 0.962) and error (MAE = 0.014 +/- 0.015 vs. 0.019 +/- 0.019), while reducing inference time by a factor of 208 (25.2 ms vs. 5 227.2 ms per image using 50 Denoising Diffusion Implicit Model (DDIM) steps). The strong clinical performance at substantially reduced computational cost enables real-time clinical use.
