Patch2Self: Denoising Diffusion MRI with Self-Supervised Learning

Abstract

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.

References in corpus (9)

• Scikit-learn: Machine Learning in Python
• Unsupervised Visual Representation Learning by Context Prediction
• Noise2Self: Blind Denoising by Self-Supervision
• Trainlets: Dictionary Learning in High Dimensions
• Non Local Spatial and Angular Matching : Enabling higher spatial resolution diffusion MRI datasets through adaptive denoising
• Linearized Kernel Dictionary Learning
• Improving Fiber Alignment in HARDI by Combining Contextual PDE Flow with Constrained Spherical Deconvolution
• Discriminative Unsupervised Feature Learning with Exemplar Convolutional Neural Networks
• Noise2Void - Learning Denoising from Single Noisy Images

Cited by in corpus (2)

• The risk of bias in denoising methods
• Denoising Improves Cross-Scanner and Cross-Protocol Test-Retest Reproducibility of Higher-Order Diffusion Metrics