3D true-phase polarity recovery with independent phase estimation using three-tier stacks based region growing (3D-TRIPS)

MAGMA. 2018 Feb;31(1):87-99. doi: 10.1007/s10334-017-0666-4. Epub 2017 Dec 7.

Abstract

Objectives: A postprocessing technique termed 3D true-phase polarity recovery with independent phase estimation using three-tier stacks based region growing (3D-TRIPS) was developed, which directly reconstructs phase-sensitive inversion-recovery images without acquisition of phase-reference images. The utility of this technique is demonstrated in myocardial late gadolinium enhancement (LGE) imaging.

Materials and methods: A data structure with three tiers of stacks was used for 3D-TRIPS to directly achieve reliable region growing for successful background-phase estimation. Fifteen patients undergoing postgadolinium 3D phase-sensitive inversion recovery (PSIR) cardiac LGE magnetic resonance imaging (MRI) were recruited, and 3D-TRIPS LGE reconstructions were compared with standard PSIR. Objective voxel-by-voxel comparison was performed. Additionally, blinded review by two radiologists compared scar visibility, clinical acceptability, voxel polarity error, or groups and blurring.

Results: 3D-TRIPS efficiently reconstructed postcontrast phase-sensitive myocardial LGE images. Objective analysis showed an average 95% voxel-by-voxel agreement between 3D-TRIPS and PSIR images. Blinded radiologist review demonstrated similar image quality between 3D-TRIPS and PSIR reconstruction.

Conclusion: 3D-TRIPS provided similar image quality to PSIR for phase-sensitive myocardial LGE MRI reconstruction. 3D-TRIPS does not require acquisition of a reference image and can therefore be used to accelerate phase-sensitive LGE imaging.

Keywords: Cardiac MRI; Late gadolinium enhancement; Myocardial infarction; Phase-sensitive imaging; Region-growing.

Publication types

  • Comparative Study

MeSH terms

  • Algorithms
  • Cardiac Imaging Techniques / methods*
  • Cardiac Imaging Techniques / statistics & numerical data
  • Computer Simulation
  • Contrast Media
  • Gadolinium
  • Humans
  • Image Interpretation, Computer-Assisted / methods
  • Imaging, Three-Dimensional / methods
  • Imaging, Three-Dimensional / statistics & numerical data
  • Magnetic Resonance Imaging / methods*
  • Magnetic Resonance Imaging / statistics & numerical data
  • Monte Carlo Method

Substances

  • Contrast Media
  • Gadolinium