Diffusion-weighted magnetic resonance imaging (DWI) is the only non-invasive method for quantifying microstructure and reconstructing white-matter pathways in the living human brain. Fluctuations from multiple sources create significant noise in DWI data which must be suppressed before subsequent microstructure analysis. We introduce a self-supervised learning method for denoising DWI data, Patch2Self, which uses the entire volume to learn a full-rank locally linear denoiser for that volume. By taking advantage of the oversampled q-space of DWI data, Patch2Self can separate structure from noise without requiring an explicit model for either. We demonstrate the effectiveness of Patch2Self via quantitative and qualitative improvements in microstructure modeling, tracking (via fiber bundle coherency) and model estimation relative to other unsupervised methods on real and simulated data.

Lasso and Multilayer Perceptron are very similar. From Figure 1, we can see that OLS and Ridge give very similar performance and the Lasso performs slightly worse. For both Ridge and Lasso, the regularization parameter alpha was set to '1.0'. The ground truth image of has been depicted on the right. Adaptive multiresolution non-local means filter for three-dimensional magnetic resonance image denoising.

Denoising is therefore a vital processing step for DWI data prior to anatomical inference.

One key difference between the proposed patch-based approach and the baseline Marchenko-Pastur is the patch-based nature of the former. Is smoother appearance of the images in figure 2 and the less noisy tractograms in figure 3 simply because the patch-based approach introduces more smoothing? That seems very likely to me. A potential big advantage of the Marchenko-Pastur is that it does not smooth so preserves detail. This is not tested in the qualitative evaluations of figures 2 and 3 or mentioned anywhere in the text.

Despite the novelty of the proposed method might be considered marginal with respect to the machine learning community, the contribution to the application field is relevant. The availability of the code represents an added value in the perspective of open science. The authors provided satisfactorily answers in the rebuttal.

Diffusion-weighted magnetic resonance imaging (DWI) is the only non-invasive method for quantifying microstructure and reconstructing white-matter pathways in the living human brain. Fluctuations from multiple sources create significant noise in DWI data which must be suppressed before subsequent microstructure analysis. We introduce a self-supervised learning method for denoising DWI data, Patch2Self, which uses the entire volume to learn a full-rank locally linear denoiser for that volume. By taking advantage of the oversampled q-space of DWI data, Patch2Self can separate structure from noise without requiring an explicit model for either. We demonstrate the effectiveness of Patch2Self via quantitative and qualitative improvements in microstructure modeling, tracking (via fiber bundle coherency) and model estimation relative to other unsupervised methods on real and simulated data.