Assessing and minimizing the effects of noise and motion in clinical DTI at 3 T

Research output: Contribution to journalArticleAcademicpeer-review

32 Citations (Scopus)

Abstract

Compared with conventional MRI, diffusion tensor imaging (DTI) is more prone to thermal noise and motion. Optimized sampling schemes have been proposed that reduce the propagation of noise. At 3 T, however, motion may play a more dominant role than noise. Although the effects of noise at 3 T are less compared with 1.5 T because of the higher signal-to-noise ratio, motion is independent of field strength and will persist. To improve the reliability of clinical DTI at 3 T, it is important to know to what extent noise and motion contribute to the uncertainties of the DTI indices. In this study, the effects of noise- and motion-related signal uncertainties are disentangled using in vivo measurements and computer simulations. For six clinically standard available sampling schemes, the reproducibility was assessed in vivo, with and without motion correction applied. Additionally, motion and noise simulations were performed to determine the relative contributions of motion and noise to the uncertainties of the mean diffusivity (MD) and fractional anisotropy (FA). It is shown that the contributions of noise and motion are of the same order of magnitude at 3 T. Similar to the propagation of noise, the propagation of motion-related signal perturbations is also influenced by the choice of sampling scheme. Sampling schemes with only six diffusion directions demonstrated a lower reproducibility compared with schemes with 15 and 32 directions and feature a positive bias for the FA in relatively isotropic tissue. Motion correction helps improving the precision and accuracy of DTI indices.

Original languageEnglish
Pages (from-to)2641-2655
Number of pages15
JournalHuman Brain Mapping
Volume30
Issue number8
DOIs
Publication statusPublished - 1 Aug 2009

Keywords

  • Clinical MRI
  • Diffusion tensor imaging
  • Gradient sampling schemes
  • Motion
  • Noise
  • Reproducibility
  • Simulations

Fingerprint Dive into the research topics of 'Assessing and minimizing the effects of noise and motion in clinical DTI at 3 T'. Together they form a unique fingerprint.

Cite this