2831
The evaluation of the metal artifact and SNR changes in multiple fat-suppression techniques：a phantom study
Xu Lulu1, Qi Liang1, Dou Weiqiang2, and Shen Yong3
1Radiology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China, 2MR Research China, GE Healthcare, Beijing, China, 3Enhanced MR application, GE Healthcare, Beijing, China
1Radiology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China, 2MR Research China, GE Healthcare, Beijing, China, 3Enhanced MR application, GE Healthcare, Beijing, China
Synopsis
Finding the most suitable fat-suppression technique when acquiring the images with titanium alloy, the phantom was studied. We found the STIR was most potent when comparing the other methods. This result may provide the reference value when we offer the patients with metallic hardware after surgery with the proper fat-suppression technique.
Introduction
T2 weighted (T2w), MRI with fat-suppression, has been widely used in routine clinical examinations. The fat-suppression techniques like short-inversion-time-inversion recovery (STIR) and FATSat are usually options to be applied accordingly. For patients with metal implanted, the application of T2w fat suppression imaging is however limited as both fat-suppression techniques are sensitive to the strong local field inhomogeneities caused by the implanted metal[1]. Iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL), as a relative novel fat-suppression technique, has been reported to be insensitive to the local field inhomogeneities[2]. However, no research has been performed to investigate these three fat-suppression techniques, i.e., STIR, FATSat and IDEAL, in terms of reducing metal artifact on T2w fat-suppression imaging. Therefore, in this study, a systematic comparison of these three techniques in T2w imaging was performed to find out the optimal technique applied in patients with metal implanted.Materials and Method
In this study, we systematically compared the performance of artifact of three kinds of fat-suppression techniques and the SNR changes with and without the screw in the phantom. It was acknowledged that the homogeneity of B0 was changed when hanging the screw. The fat resonance frequency was variety by the B0, so it was challenging to suppress it when applying the FATSat technique, thus FATSat had the highest R1, R2 and the artifact area. IDEAL fat showed an excellent performance than FATSat because it can give a correct phase of fat and water in somewhat by three equations. Finally, we believed that the fat’s T1 value changed little than the variety of frequency, so the area artifact in STIR was the smallest. As shown in the results, STIR-SNR was increased in phantom images with screws than without, while the opposite pattern was observed for FATSat-SNR and IDEAL-SNR. The phenomenon can be explained that IDEAL and FATSat were all based on the Chemical-shift Techniques, and the inhomogeneity of B0 made the T2 relaxation time faster, which led to lower signal, and then the SNR did so.Results
For R1, the most significant value was shown in FATSat, followed by IDEAL fat and STIR (p<0.01; Fig.1.a). The smallest artifact area was founded by using the STIR, while the biggest was saw in the FATSat. Similar patterns were also found for R2 (Fig.1.b-c). In addition, the SNR values for ROIs acquired at phantom1 and phantom2 were also respectively compared for each fat-saturation technique. As shown in Fig2.d-f, an increased SNR level was observed in phantom with metal than without for STIR case and opposite patterns were however shown for FATSat and IDEAL fat cases (all p < 0.01).Discussion
In this study, we systematically compared the performance of artifact of three kinds of fat-suppression techniques and the SNR changes with and without the screw in the phantom. It was acknowledged that the homogeneity of B0 was changed when hanging the screw. The fat resonance frequency was variety by the B0, so it was challenging to suppress it when applying the FATSat technique; thus FATSat had the highest R1, R2 and the artifact area. IDEAL fat showed an excellent performance than FATSat because it can give a correct phase of fat and water in somewhat by three equations. Finally, we believed that the fat’s T1 value changed little than the variety of frequency, so the area artifact in STIR was the smallest. As shown in the results, STIR-SNR was increased in phantom images with screws than without, while the opposite pattern was observed for FATSat-SNR and IDEAL-SNR. The phenomenon can be explained that the titanium alloy was paramagnetic material so that it may shorten the fat T1 value, then the null point in STIR was changed, whereas IDEAL and FATSat were all based on the Chemical-shift Techniques, and the inhomogeneity of B0 made the T2 relaxation time faster, which led to lower signal, and then the SNR did so.Conclusion
In conclusion, among the three fat-suppression techniques tested in this study, STIR showed the most robust performance on the artifact reduction.Acknowledgements
No acknowledgement found.References
[1]Del Grande F, et al. Radiographs, 2014, 34 (1): 217-233.
[2]Cha J G et al. Radiology, 2011, 259 (3): 885-893.
[3]ASTM F2119-01, Standard Test Method for Evaluation of MR Image Artifacts from Passive Implants, ASTM International, West Conshohocken, PA, 2001, www.astm.org
[4]National Electrical Manufacturers Association. NEMA Standards Publication MS 1-2008 Determination of Signal to Noise Ratio (SNR) in diagnostic magnetic resonance images; 2008.
Figures
Note：BW= half-bandwidth in GE
Fig.1 a and b show the results of R1 and R2. c displays the difference of artifact area with three kinds of fat-suppression.
Fig.2
The labels above originate from phantom1. The labels below(’) stand for the
parameters from phantom2. ROI=region of interest, SD=standard deviation. Background
intensity = mean level of signal intensities of background. The Y-axis represents the signal strength or SD and The X-axis means the serial number of the screw.