T2* Measurement bias due to concomitant gradient fields

Lorne W. Hofstetter*, Glen Morrell, Joshua Kaggie, Daniel Kim, Kristi Carlston, Vivian S. Lee

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

6 Scopus citations


Purpose: To demonstrate that concomitant magnetic fields can cause significant spatially dependent biases in T* 2 relaxometry measurements with implications for clinical applications such as BOLD and dynamic susceptibility contrast-enhanced MRI. Theory and Methods: After developing a theoretical framework for intravoxel dephasing and signal loss from concomitant magnetic fields, this framework and the effect of concomitant fields on T* 2 are validated with phantom experiments and numerical simulation. In lower leg and renal T* 2 mapping, we quantify measurement bias for imaging protocols with high gradient amplitude multiecho readouts, comparable to those used in clinical applications. Results: Concordance between phantom experiment and numerical simulation validate the theoretical framework. Changes in T* 2 measured in the lower leg and kidney varied by up to 15% and 35%, respectively, as a result of concomitant gradient effects when compared with the control measurements. Conclusion: Concomitant magnetic fields produced by imaging gradient coils can cause clinically significant T* 2 mapping errors when high amplitude, long duration gradient waveforms are used. While we have shown that measurement biases can be quite large, modification of imaging parameters can potentially reduce concomitant field-induced measurement errors to acceptable levels. Magn Reson Med 77:1562–1572, 2017.

Original languageEnglish (US)
Pages (from-to)1562-1572
Number of pages11
JournalMagnetic resonance in medicine
Issue number4
StatePublished - Apr 1 2017


  • BOLD
  • T2*
  • concomitant gradient fields
  • iron quantification
  • renal-BOLD

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging


Dive into the research topics of 'T2* Measurement bias due to concomitant gradient fields'. Together they form a unique fingerprint.

Cite this