Flow-induced artifacts in two-point Dixon MRI: Incidence, severity and potential diagnostic pitfalls.
Tilman Schubert1, Peter Bannas2, Samir Sharma3, Sonja Kinner1, Mahdi Rahimi3, Frank Korosec3, and Scott Reeder1

1Radiology, University of Wisconsin Madison, Madison, WI, United States, 2Radiology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany, 3Medical Physics, University of Wisconsin Madison, Madison, WI, United States

Synopsis

Chemical shift based two-point “Dixon” MRI with bipolar readout gradients may produce flow induced fat-water misallocation artifacts. These artifacts have the potential to mimic intravascular thrombus. We reviewed 100 cases of two-point body MRI exams to characterize the incidence, location and severity of these artifacts. Artifacts appeared in 46% of the cases, with severe artifacts in 20% and mild artifacts in 26% of the patients. Given this high number of potentially thrombus-mimicking, flow-induced artifacts, radiologists should be aware of this potential pitfall when using two-point fat-water separation methods.

Purpose

Chemical shift based two-point “Dixon” fat-water separation magnetic resonance imaging (MRI) techniques provide robust fat saturation and have recently been applied to MR angiography (MRA) [1]. However, the acquisition of two echoes increases scan time. In order to minimize scan time, the use of bipolar readout gradients to sample both echoes in the same TR is commonly applied. However, with a bipolar readout acquisition, the first moments of the two echoes are different, leading to phase differences between the two echoes described by the following equation $$\triangle\phi=\gamma\times \triangle M1\times\upsilon$$where g is the gyromagnetic ratio of 1H, DM1 is the difference in the first moment of the readout gradient between echoes, and is the velocity of a moving spin along the readout direction. These additional flow-induced phase shifts confound the separation of the fat and water signals, and can lead to misallocation of the phase information of spins flowing in the readout direction [2]. The purpose of this study was to characterize the incidence, location and severity of these artifacts in body MRI.

Methods

Two-point “Dixon” coronal and axial body-MRI images (LAVA Flex, GE Healthcare, Waukesha, WI, USA) acquired in 100 consecutive patients were retrospectively evaluated for the presence and location of flow-induced artifacts in arteries and veins. Artifacts were graded on a 3-point scale (none, mild, severe) by two radiologists independently. Inter-reader agreement was evaluated with kappa statistics.

Results

Intravascular fat-water misallocation artifacts were present in 46 patients (46%) with severe artifacts in 20 (20%) and mild artifacts in 26 (26%) patients. Artifacts were present on coronal images in 40 patients and on axial images in 16 patients, totaling 70 artifacts. 31/70 artifacts (44%) were graded as severe and 39/70 (56%) as mild (30 coronal, 9 axial). Misallocation of fat and water was apparent in all datasets with severe artifacts whereas variable signal intensity changes in water and fat images were observed in mild artifacts. Inter-reader agreement was good for artifacts appearing in coronal imaging (κ=0.7) and fair for artifact appearance in axial imaging (κ=0.24). Figure 1 shows a mild misallocation artifact in the IVC and the Aorta with varying signal intensities in the fat-only image. Figure 2 shows a typical example of a severe, thrombus-mimicking misallocation of fat signal into the water-only image and vice versa (arrows) in the IVC. Figure 3 shows an example of a severe misallocation artifact in both renal arteries.

Discussion

Artifact location and appearance are consistent with previous reports [2], The artifacts were observed particularly in the coronal slice orientation and in vessels with flow in the same direction as the readout gradient. Misallocation artifacts appeared more frequently in the venous system with a predilection to occur in the intrahepatic IVC. Our study has important clinical implications as imaging artifacts in MRI are known to mimic thrombus [3]. However, severe flow induced misallocation artifacts can be correctly diagnosed as artifacts, if the fat images are consistently reviewed together with the water images, as the misallocation of water signal into the fat image can be clearly visualized.

Conclusion

Flow-induced misallocation artifacts with the potential to mimic intravascular thrombus appear in almost 50% of patients with two-point Dixon MR-exams. Radiologists should be aware of this potential pitfall when using two-point Dixon fat-water separation methods.

Acknowledgements

The authors wish to acknowledge support from the NIH (UL1TR00427, R01 DK083380, R01 DK088925, R01 DK100651, K24 DK102595), as well as GE Healthcare.

References

[1] Leiner et al, Eur Radiol. 2013, [2] Rahimi et al, Magn Reson Med. 2015 [3] Bannas et al, European Radiol 2014

Figures

Example of a mild misallocation artifact in the water-only image (A) in the IVC (arrow) and the Aorta (arrowhead) with varying signal intensities in the fat-only image (B).

Example of a severe, thrombus-mimicking misallocation of fat signal into the water-only image (A) and vice versa (arrows) in the IVC.

Example of a severe misallocation artifact in both renal arteries. Water-only (upper) and fat-only image (lower).



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4313