Comparison of reconstruction and acquisition choices for quantitative T2* maps and synthetic contrasts

Riikka Ruuth, Linda Kuusela, Teemu Mäkelä, Susanna Melkas, Antti Korvenoja

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Aim and scope: A Gradient Echo Plural Contrast Imaging technique (GEPCI) is a post-processing method, which can be used to obtain quantitative T2* values and generate multiple synthetic contrasts from a single acquisition. However, scan duration and image reconstruction from k-space data present challenges in a clinical workflow. This study aimed at optimizing image reconstruction and acquisition duration to facilitate a post-processing method for synthetic image contrast creation in clinical settings.

Materials and methods: This study consists of tests using the American College of Radiology (ACR) image quality phantom, two healthy volunteers, four mild traumatic brain injury patients and four small vessel disease patients. The measurements were carried out on a 3.0 T scanner with multiple echo times. Reconstruction from k-space data and DICOM data with two different coil-channel combination modes were investigated. Partial Fourier techniques were tested to optimize the scanning time.

Conclusions: Sum of squares coil-channel combination produced artifacts in phase images, but images created with adaptive combination were artifact-free. The voxel-wise median signed difference of T2* between the vendor's adaptive channel combination and k-space reconstruction modes was 2.9 +/- 0.7 ms for white matter and 4.5 +/- 0.6 ms for gray matter. Relative white matter/gray matter contrast of all synthetic images and contrast-to-noise ratio of synthetic T1-weighted images were almost equal between reconstruction modes. Our results indicate that synthetic contrasts can be generated from the vendor's DICOM data with the adaptive combination mode without affecting the quantitative T2* values or white matter/gray matter contrast.

Original languageEnglish
JournalEuropean journal of radiology open
Volume6
Pages (from-to)42-48
Number of pages7
ISSN2352-0477
DOIs
Publication statusPublished - 2019
MoE publication typeA1 Journal article-refereed

Fields of Science

  • 3112 Neurosciences
  • 3124 Neurology and psychiatry
  • 3126 Surgery, anesthesiology, intensive care, radiology
  • 114 Physical sciences
  • Magnetic resonance imaging
  • Image quality
  • Quantitative MRI
  • MRI reconstruction
  • T2* mapping
  • Synthetic contrasts
  • HUMAN BRAIN
  • ROBUST
  • OPTIMIZATION
  • REGISTRATION
  • RELAXATION
  • DISEASE
  • PHASE

Cite this