Characterization and Correction of Time-Varying Eddy Currents for Diffusion MRI
2021
Diffusion MRI (dMRI) generally suffers from eddy currents induced by strong diffusion gradients, which introduce artefacts that can impair subsequent diffusion metric analysis. Existing correction techniques assume that eddy currents do not decay during acquisition, which is generally effective for traditional Pulsed Gradient Spin Echo (PGSE) diffusion encoding. However, these methods do not necessarily apply to advanced forms of dMRI that require substantial gradient slewing, such as Oscillating Gradient Spin Echo (OGSE). In this work, dynamic field monitoring was used to characterize diffusion gradient induced eddy currents for both PGSE and OGSE to assess the applicability of these typical assumptions. The performance of an in-house algorithm (TVEDDY) that models eddy current decay was evaluated by correcting in-vivo PGSE and OGSE brain images and comparing correction quality with conventional methods using mean-squared error (MSE) between diffusion weighted images acquired with opposite polarity diffusion gradients. As a ground truth comparison, spatially varying field dynamics up to third order in space were used in a model-based iterative reconstruction to eliminate distortions. Time-varying eddy currents were observed for OGSE, which introduced blurring along the phase-encode direction that was not reduced using the traditional approach, but was diminished considerably with TVEDDY and model-based reconstruction. No MSE difference was observed between the conventional approach and TVEDDY for PGSE, but for OGSE TVEDDY resulted in significantly lower MSE. The field-monitoring-informed model-based reconstruction had the lowest MSE for both PGSE and OGSE. Accordingly, this work establishes that it is possible to estimate time-varying eddy currents from the diffusion data itself, which provides substantial image quality improvements for gradient-intensive dMRI acquisitions like OGSE.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
20
References
0
Citations
NaN
KQI