Field inhomogeneity correction for gradient echo myelin water fraction imaging

Magn Reson Med. 2017 Jul;78(1):49-57. doi: 10.1002/mrm.26334. Epub 2016 Jul 15.

Abstract

Purpose: Recently, the multi-echo gradient echo (MGRE) sequence has been proposed for multicomponent T2* (MC T2*) based myelin water fraction (MWF) mapping. This approach has appeal because it can provide fast whole-brain coverage, has low specific absorption rate, and short echo spacing. However, the MGRE signal requires correction for accurate MWF mapping, because of its sensitivity to magnetic field inhomogeneities (ΔB0 ). We propose a ΔB0 correction method for 2D MGRE data obtained for MWF mapping.

Theory and methods: Latter-echo MGRE data were fit to estimate B0 gradients in the slice-select direction ( Gz). The decay signal was corrected for the effects of Gz, and MC T2* analysis was performed using nonnegative least-squares fitting. The method was evaluated using simulations and its performance demonstrated in healthy volunteers.

Results: Simulations showed that MWFs are significantly biased in the presence of Gz and that our correction method leads to accurate MWF estimates. In vivo MWF maps obtained from corrected data showed recovery of MWF estimates in areas of high ΔB0, and overall good agreement with literature values obtained with the reference MC T2-based method.

Conclusion: A new algorithm was presented for ΔB0 correction of 2D MGRE echo data acquired for MWF imaging. Simulations and in vivo data showed an improvement in MWF estimates. Magn Reson Med 78:49-57, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Keywords: field inhomogeneities; gradient echo imaging; multicomponent T2*; myelin water.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Artifacts*
  • Body Water / chemistry*
  • Body Water / diagnostic imaging
  • Corpus Callosum / chemistry*
  • Diffusion Tensor Imaging / methods*
  • Female
  • Humans
  • Image Enhancement / methods
  • Male
  • Molecular Imaging / methods*
  • Myelin Sheath / chemistry*
  • Nerve Fibers, Myelinated / chemistry*
  • Reproducibility of Results
  • Sensitivity and Specificity
  • Signal Processing, Computer-Assisted

Grants and funding