Patch2Self: Denoising Diffusion MRI with Self-Supervised Learning
This addresses noise suppression in DWI data for brain microstructure analysis, which is incremental as it builds on existing unsupervised denoising approaches.
The paper tackles denoising diffusion-weighted MRI data by introducing Patch2Self, a self-supervised learning method that uses the oversampled q-space to separate structure from noise without explicit models, resulting in quantitative and qualitative improvements in microstructure modeling, tracking, and model estimation compared to other unsupervised methods.
Diffusion-weighted magnetic resonance imaging (DWI) is the only noninvasive method for quantifying microstructure and reconstructing white-matter pathways in the living human brain. Fluctuations from multiple sources create significant additive 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.