Reduction of Distortion Artifacts in Brain MRI Using a Field Map-based Correction Technique in Diffusion-weighted Imaging : A Prospective Study

Clin Neuroradiol. 2024 Mar;34(1):85-91. doi: 10.1007/s00062-023-01338-3. Epub 2023 Aug 28.

Abstract

Purpose: The aim of this study was to evaluate the image quality and feasibility of a field map-based technique to correct for susceptibility-induced geometric distortions which are typical for diffusion EPI brain imaging.

Methods: We prospectively included 52 patients during clinical routine in this single-center study. All scans were performed on a 3T MRI. Patients' indications for MRI mainly consisted of suspected stroke due to the clinical presentation. For the morphological comparison of the corrected and uncorrected EPI diffusion, three experienced radiologists assessed the image quality of the sequences in a blinded and randomized fashion using a Likert scale (1 being poor; 5 being excellent). To ensure comparability of the two methods, an additional quantitative analysis of the apparent diffusion coefficient (ADC) was performed.

Results: Corrected EPI diffusion was rated significantly superior in all the selected categories: overall level of artifacts (p < 0.001), degree of distortion at the frontal, temporal, occipital and brainstem levels (p < 0.001), conspicuousness of ischemic lesions (p < 0.001), image quality (p < 0.001), naturality (p < 0.001), contrast (p < 0.001), and diagnostic confidence (p < 0.001).

Conclusion: Corrected EPI diffusion offers a significant reduction of geometric distortion in all evaluated brain regions and an improved conspicuousness of ischemic lesions. Image quality, overall artifacts, naturality, contrast and diagnostic confidence were also rated superior in comparison to uncorrected EPI diffusion.

Keywords: Diffusion-weighted imaging; Field map correction; Geometric distortion; Image quality; Stroke imaging.

MeSH terms

  • Artifacts*
  • Brain / diagnostic imaging
  • Diffusion Magnetic Resonance Imaging / methods
  • Echo-Planar Imaging* / methods
  • Humans
  • Magnetic Resonance Imaging
  • Neuroimaging
  • Prospective Studies
  • Reproducibility of Results
  • Sensitivity and Specificity