B0 Eddy Current Correction for Spiral MRI
Ryan K Robison1, Dinghui Wang1, Zhiqiang Li1, and James G Pipe1

1Imaging Research, Barrow Neurological Institute, Phoenix, AZ, United States

Synopsis

Eddy currents are a common source of artifacts in Spiral MRI. Eddy currents that effect the k-space trajectory are often the focus of eddy current correction. However, the spatially uniform but time-varying B0 eddy currents can also be a subtle but important source of artifacts in spiral images. This work demonstrates the improvement in image quality that can result from measuring and correcting the phase produced by B0 eddy currents in spiral MRI.

Introduction

B0 eddy currents are a subtle but important source of artifact in spiral MRI. As the gradient amplitude changes, B0 eddy currents produce time dependent B0 field shifts on the order of hundreds of hertz (see figure 1) and, consequently, time dependent phase shifts in the measured data. Previously, it has been shown that B0 eddy currents cause artifacts in 3D projection reconstruction sequences1 and blurriness in combined spiral-in/out images2. This work demonstrates the importance of accounting for B0 eddy currents in all spiral sequences.

Methods

Sagittal spiral-out MPRAGE data and axial spiral-in/out TSE data were acquired in a healthy volunteer. The method of Brodsky et al.1 was used to measure the phase produced by B0 eddy currents during the spiral acquisitions. Data were corrected by applying the conjugate of the measured B0 eddy current induced phase to the spiral imaging data. All data were processed and reconstructed in GPI3 using standard gridding reconstruction. The measured B0 eddy current induced phase was gridded to yield maps of the B0 phase distribution over k-space.

Results

The instantaneous slew rate and resultant measured f0 shift due to B0 eddy currents are shown in figure 1 for a single spiral interleave. As expected, the B0 eddy currents depend upon the rate at which the gradient amplitude is changing. Figure 2 shows gridded B0 eddy current phase maps. The B0 eddy currents produce a roughly linear phase across k-space in the axial case, and a non-linear distribution of B0 eddy current phase in the sagittal case. The resultant image artifacts from the B0 eddy currents are highlighted in figure 3. The linear B0 eddy current phase across k-space, in the axial case, causes opposite shifts in image space for the spiral-in and spiral-out portions of the acquisition and, consequently, blurriness in the combined spiral-in/out TSE image. The non-linear distribution of B0 eddy current phase in the sagittal case instead produces signal pileups and voids somewhat reminiscent of chemical shift artifacts. The artifacts are favorably improved in both cases by correcting for the B0 eddy current phase.

Discussion and Conclusion

This work has demonstrated that B0 eddy currents can produce substantial artifacts in any spiral acquisition. These artifacts can be subtle in nature and difficult to distinguish from blurring, k-space trajectory errors, and chemical shift. It should be noted that the results shown here are for a single scanner. The size and distribution of B0 eddy currents are expected to vary from scanner to scanner. Nevertheless, these effects should be taken into consideration for any high resolution spiral imaging.

Acknowledgements

This work was funded in part by Philips Healthcare

References

1. Brodsky EK, Klaers JL, Samsonov AA, Kijowski R, and Block WF. "Rapid Measurement and Correction of Phase Errors From B0 Eddy Currents: Impact on Image Quality for Non-Cartesian Imaging", MRM (2013), 69: 509-515.

2. Robison RK, Li Z, Schar M, Pipe JG. "Correction of B0 Phase Errors for Spiral-In/Spiral-Out Acquisitions", Proceedings of the Joint Annual Meeting ISMRM-ESMRMB 2014, Milan, Italy, Abstract # 1620.

3. Zwart NZ and Pipe JG. "Graphical programming interface: A development environment for MRI methods", MRM (2015), 74:1449-1460.

Figures

Figure 1: Designed instantaneous slew rate (a) and measured B0 eddy current induced frequency shift (b) for one spiral interleave

Figure 2: Measured B0 eddy current induced phase gridded onto k-space for an axial (a) and sagittal (b) spiral acquisition. The phase scale is given in the color bar around the images. A cross-section of the phase map is also plotted for each case to better demonstrate the distribution of the phase across k-space.

Figure 3: Reconstructed spiral-in/out TSE (a,b) and spiral-out MPRAGE (c,d) images. Data were reconstructed without (a,c) and with (b,d) correction of the B0 eddy current induced phase. Zoomed images (inset) highlight the value of the B0 eddy current correction.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0051