Self-calibrated multiple-echo acquisition with radial trajectories using the conjugate gradient method (SMART-CG)

Youngkyoo Jung, Alexey A. Samsonov, Mark Bydder, Walter F. Block

Research output: Contribution to journalArticle

Abstract

Purpose: To remove phase inconsistencies between multiple echoes, an algorithm using a radial acquisition to provide inherent phase and magnitude information for self correction was developed. The information also allows simultaneous support for parallel imaging for multiple coil acquisitions. Materials and Methods: Without a separate field map acquisition, a phase estimate from each echo in multiple echo train was generated. When using a multiple channel coil, magnitude and phase estimates from each echo provide in vivo coil sensitivities. An algorithm based on the conjugate gradient method uses these estimates to simultaneously remove phase inconsistencies between echoes, and in the case of multiple coil acquisition, simultaneously provides parallel imaging benefits. The algorithm is demonstrated on single channel, multiple channel, and undersampled data. Results: Substantial image quality improvements were demonstrated. Signal dropouts were completely removed and undersampling artifacts were well suppressed. Conclusion: The suggested algorithm is able to remove phase cancellation and undersampling artifacts simultaneously and to improve image quality of multiecho radial imaging, the important technique for fast three-dimensional MRI data acquisition.

Original languageEnglish (US)
Pages (from-to)980-987
Number of pages8
JournalJournal of Magnetic Resonance Imaging
Volume33
Issue number4
DOIs
StatePublished - Apr 1 2011

Keywords

  • Multi-echo acquisition
  • Off-resonance correction
  • Radial imaging
  • Sensitivity encoding

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Fingerprint Dive into the research topics of 'Self-calibrated multiple-echo acquisition with radial trajectories using the conjugate gradient method (SMART-CG)'. Together they form a unique fingerprint.

  • Cite this