Track trials, findings, and clinical evidence as it lands.
Search papers, labs, and topics across Lattice.
Standard self-supervised blind-spot denoisers like Noise2Void produce severe parametric mapping bias on magnitude cardiac MRI because their mean-squared-error objective assumes zero-mean Gaussian noise rather than a signal-dependent Rician distribution. To resolve this, the authors reformulate self-supervised blind-spot training as maximum likelihood estimation under an explicit Rician noise model for short-axis $T2$- and $T1\rho$-weighted sequences. This formulation eliminates estimation bias and recovers quantitative parametric maps with performance competitive with fully supervised baselines, entirely without clean target data.
Standard self-supervised denoising silently corrupts quantitative MRI biomarkers by assuming Gaussian noise, but swapping MSE for an exact Rician likelihood matches fully supervised quality without needing clean ground truth.
Magnetic resonance imaging involves an inherent trade-off among spatial resolution, acquisition time, and noise. This trade-off contributes to long scan times and high cost. Deep learning has improved image denoising, but cardiac MRI remains difficult because high-resolution, rapid acquisitions generally lack corresponding low-noise ground truth. Self-supervised denoising offers a potential solution by learning from noisy image pairs or even single noisy acquisitions. However, we show that Noise2Void-style blind-spot denoising, which uses a mean squared error (MSE) loss and assumes zero-mean, independent and identically distributed (i.i.d.) noise, is poorly suited to MR magnitude images. When applied to short-axis $T2$-weighted and $T1\rho$-weighted cardiac MRI with synthetic Rician noise, it produces biased denoised images and biased parametric maps of $T2$ and $T1\rho$. To address this limitation, we formulate self-supervised denoising as maximum likelihood estimation under a known Rician noise model. This yields unbiased denoisers that are competitive with supervised baselines.