4139
Optimized Acceleration Factor in Phase Measurement of Brain Deep Veins using SWI with Compressed Sensing1First Affiliated Hospital of Dalian Medical University, Dalian, China, 2Philips Healthcare, Dalian, China
Synopsis
The MR scan time can be decreased using Compressed sensing with a favorable image quality. The study aimed to quantitative measure the oxygen concentration using Susceptibility Weighted Imaging (SWI) images with the different acceleration factors (AF), and explore the optimized AF of compressed sensing based on quantitative assessment of the image quality. Combined with the SNR and CNR, when the AF is 4, the scanning time can be reduced by 71.39%. For the phase value of deep veins, the AF can reach to 6 and the scan time is reduced to 1min8s without image quality change obviously.
Introduction
Susceptibility Weighted Imaging (SWI) is extremely sensitive to small magnetic field inhomogeneity, which mainly comes from deoxyhemoglobin, methemoglobin, hemosiderin and ferritin etc[1]. Based on SWI, oxygen content can be measured in the deep veins, which is useful in the diagnosis of many diseases (such as various hemorrhagic lesions, abnormal venous vascular lesions, tumors and degenerative diseases)[2]. However, the SWI sequence is time-consuming and is easy to be affected by motion artifacts, which leads to the decrease of image resolution and the image quality.Compression sensing (CS) can shorten the time required for signal acquisition and reduce the calculation amount, and to some extent, meeting the requirements of maintaining the reconstruction quality of the original signal[3]. However, the image quality will be decreased with AF increased. In this study, we aim to explore the optimized AF in quantitative measuring oxygen content in the deep veins based on SWI with CS.
Materials and Methods
The study recruited 20 health volunteers (mean age:51.9 ± 14.97, range:27 - 84 years, 13 females and 7 males). All MR images were acquired using a 3.0T MRI scanner (Ingenia CX, Philips Healthcare, the Netherlands). The MR protocol included routine SWI sequence without acceleration factor (AF = 0) and with CS of different acceleration factors (AF = 2, 4, 6, 8, 10). The detailed scan parameters are shown in Table 1. Phase images of SWI were post-processed and measured using the Signal Processing in Nuclear Magnetic Resonance (SPIN) software. The ROIs were draw manually in five brain veins including the septal vein (SV), internal cerebral vein (ICV), superior thalamostriate vein (STV), basal vein (BV), dentate nucleus vein (DNV) of lateral ventricle on phase images (Figure 1). White matter (WM) and cerebrospinal fluid (CSF) were also measured as the reference. The mean and variance of phase values among above ROIs were recorded with different AFs. The SNR and CNR are calculated by the following two formulas: (SNRWM-CSF)=SIWM/SDCSF，(CNRdeep veins-WM)=(SIdeep veins－SDWM)/SDWM. P<0.05 considered statistically significant.Result
The scan time of SWI sequence with six acceleration factors (0, 2, 4, 6, 8, 10) are 5min53s, 3min21s, 1min41s, 1min8s, 51s and 41.1s. From figure 2, we can see that the resolution reduction of the veins with different AFs. As for SNR, there was no significant difference in SNR between WM and CSF in all of AFs (Figure 3). As for the CNR, there was no significant difference of deep veins compared with WM in AF of 4. When the AF was 6, the first difference appeared in the right SV, the left ICV, the right STV and the right BV. As factors increased from 8 to 10, there were differences in more and more brain veins, such as the right ICV, the left STV, the left BV, the bilateral DNVs (Figure 3). As for the phase value of the veins, there were differences with AF of 8 in deep veins of the left SV, the left ICV, the right STV and the right BV compared with AF of 0. And there were no significant differences in the right ICV, the left STV, the left BV, the DNVs on both sides(Figure 4).Discussion
In this study, we found there are the differences of SWI image quality in deep veins with different acceleration factors, which may quantitatively reflect the oxygen concentration. There was no significant difference in SNR between WM and CSF, but when the AF increased to 6, there are significant differences of CNR in some veins and white matter compared with those with AF of 0. So the optimized AF is 4. However, as far as the phase value is concerned, the AF of 6 is the optimized acceleration factor. Therefore, the optimal acceleration factor is 6 based on venous measurement, and the scan time is reduced to 1min8s.The susceptibility of the oxygenated hemoglobin is very small relative to the magnetic susceptibility of the surrounding brain tissue, and deoxyhemoglobin is paramagnetic relative to the surrounding brain tissue. Therefore, changes in oxygen saturation can cause phase differences between the veins and surrounding tissues[4]. By measuring the phase difference between the small vein and the surrounding tissue, the information of venous oxygen content can be obtained[5].
Conclusion
In this study, we found there are no obvious differences on the display of each vein in the brain of AF of 6 in SWI image with CS. Venous oxygen concentration is the basis of SWI vascular imaging, and venous phase values can reflect venous blood oxygen levels. It has positive significance for clinical application to find ischemic and hypoxic lesions in blood vessels and surrounding tissues of critical patients.Acknowledgements
No acknowledgement found.References
[1] Liu S, Buch S, Chen Y, Haacke EM, et al. Susceptibility-weighted imaging: current status and future directions. NMR Biomed, 2017, 30(4)
[2] Grabner G, Kiesel B, Wöhrer A, et al. Local image variance of 7 Tesla SWI is a new technique for preoperative characterization of diffusely infiltrating gliomas: correlation with tumour grade and IDH1 mutational status. Eur Radiol, 2017, 27(4): 1556-1567.
[3] Mönch S, Sollmann N, Hock A, Hedderich DM, et al. Magnetic Resonance Imaging of the Brain Using Compressed Sensing – Quality Assessment in Daily Clinical Routine. Clin Neuroradiol. 2020 Jun, 30(2):279-286
[4] Schaller B, Graf R. Cerebral venous infarction: the pathophysiological concept. Cerebrovasc Dis, 2004, 18(3): 179-188.
[5] Schmierer K, Parkes HG, So PW, et al. High field (9.4 Tesla) magnetic resonance imaging of cortical grey matter lesions in multiple sclerosis. Brain, 2010, 133(Pt 3): 858-867.
Figures
