Denoising of diffusion MRI using random matrix theory

Neuroimage. 2016 Nov 15:142:394-406. doi: 10.1016/j.neuroimage.2016.08.016. Epub 2016 Aug 11.


We introduce and evaluate a post-processing technique for fast denoising of diffusion-weighted MR images. By exploiting the intrinsic redundancy in diffusion MRI using universal properties of the eigenspectrum of random covariance matrices, we remove noise-only principal components, thereby enabling signal-to-noise ratio enhancements. This yields parameter maps of improved quality for visual, quantitative, and statistical interpretation. By studying statistics of residuals, we demonstrate that the technique suppresses local signal fluctuations that solely originate from thermal noise rather than from other sources such as anatomical detail. Furthermore, we achieve improved precision in the estimation of diffusion parameters and fiber orientations in the human brain without compromising the accuracy and spatial resolution.

Keywords: Accuracy; Marchenko-Pastur distribution; PCA; Precision.

MeSH terms

  • Data Interpretation, Statistical*
  • Diffusion Magnetic Resonance Imaging / methods*
  • Humans
  • Image Processing, Computer-Assisted / methods*
  • Principal Component Analysis / methods*
  • White Matter / diagnostic imaging*