Skip to main navigation Skip to search Skip to main content

Susceptibility-induced internal gradients reveal axon morphology and cause anisotropic effects in the diffusion-weighted MRI signal

  • S. Winther*
  • , H. Lundell
  • , J. Rafael-Patiño
  • , M. Andersson
  • , J-P. Thiran
  • , T. B. Dyrby
  • *Corresponding author for this work
  • Swiss Federal Institute of Technology Lausanne
  • Copenhagen University Hospital Amager and Hvidovre

Research output: Contribution to journalJournal articleResearchpeer-review

24 Downloads (Orbit)

Abstract

Diffusion-weighted MRI is our most promising method for estimating microscopic tissue morphology in vivo. The signal acquisition is based on scanner-generated external magnetic gradients. However, it will also be affected by susceptibility-induced internal magnetic gradients caused by interactions between the tissue and the static magnetic field of the scanner. With 3D in silico experiments, we show how internal gradients cause morphology-, compartment-, and orientation-dependence of spin-echo and pulsed-gradient spin-echo experiments in myelinated axons. These effects surpass those observed with previous 2D modelling corresponding to straight cylinders. For an ex vivo monkey brain, we observe the orientation-dependence generated only when including non-circular cross-sections in the in silico morphological configurations, and find orientation-dependent deviation of up to 17% for diffusion tensor metrics. Interestingly, we find that the orientation-dependence not only biases the signal across different brain regions, but also carries a sensitivity to the morphology of axonal cross-sections which is not attainable by the idealised theoretical diffusion-weighted MRI signal.
Original languageEnglish
Article number29636
JournalScientific Reports
Volume14
Number of pages18
ISSN2045-2322
DOIs
Publication statusPublished - 2024

Fingerprint

Dive into the research topics of 'Susceptibility-induced internal gradients reveal axon morphology and cause anisotropic effects in the diffusion-weighted MRI signal'. Together they form a unique fingerprint.

Cite this