Supplementary Data for Liver MR Imaging Including MRE Acquisition/Measurement

Supplementary data for liver MR imaging including MRE acquisition/Measurement

1. Detailed MR imaging protocol

All patients fasted for at least eight hours prior to the MR examination. MR images were obtained from the level of liver dome to right renal hilum at axial scan, and from the liver dome to the iliac crest at coronal scan. MR images were composed of a respiratory-triggered T2-weighted fast spin echo sequence, a T2-weighted single-shot fast spin-echo (SSFSE) sequence, a breath-hold T1-weighted, dual-echo (in-phase and opposed-phase) spoiled gradient recalled echo (GRE) sequence, and a T2*- weighted, fast gradient echo sequence. Three-dimensional (3D), fat-saturated, spoiled, gradient-echo sequences (liver acquisition with volume acceleration; LAVA, GE Medical Systems) were obtained both before and after IV bolus administration of gadoxetic acid (Primovist®, Bayer Schering Pharma AG, Berlin, Germany) at a dose of 0.025 mmol/kg (0.1 mL/kg body weight) at a rate of 1mL/s and immediately followed by a 30-mL saline flush through an antecubital venous catheter using a power injector (Stellant Dual, Siemens Medical Solutions). Dynamic MR imaging included the following phases: hepatic arterial phase, portal venous phase, transitional phase, and hepatobiliary phase. Scanning delay times after contrast injection determined by real-time MRI fluoroscopic monitoring were as follows: arterial phase, 7 seconds after contrast media arrival at a distal thoracic aorta; portal venous phase, 60 seconds after contrast injection; transitional phase, 3 minutes after contrast injection; and hepatobiliary phase, 20 minutes after contrast medium injection.

2. Sequence parameters

Sequence / TR/TE (ms) / Flip Angle / Echo Train Length / FOV (mm) / Matrix / Section Thickness / Intersection Gap
Respiratory-triggered T2 FSE / 8000-10000/103 / 90 º / 16 / 380 x 380 / 448 x 256 / 7mm / 0
Breath-hold T2 SSFSE / 850/160 / 90 º / 1 / 380 x 380 / 320 x 192 / 7mm / 0
Breath-hold T1 spoiled GRE / 6.6/4.4 (in)
6.6/2.1 (oppsed) / 12 º / 1 / 380 x 380 / 320 x 224 / 4.8 mm / 1.2mm
T2*- weighted, FGRE / 100/15 / 30 º / 1 / 380 x 380 / 320 x 288 / 7mm / 0
Breath-hold T1 3D-LAVA / 4.6/2.3 / 12 / 1 / 380 x 380 / 320 x224 / 4.8 mm / 1.2mm

Note.━ TR/TE=relaxation time/echo time, FOV=field of view, FSE=fast spin echo, SSFSE=single shot fast spin echo, GRE=gradient recalled echo, 3D=three dimensional, T1 3D-LAVA = T1-weighted three-dimensional liver acquisition with volume acceleration

3. Techniques of MR elastography

The MR elastography parameters were as follows: 1.5T GE Signa Excite scanner software version 16M4, repetition time/echo time, 1333.8/51.2ms; excitation flip angle, 90°; refocusing flip angle, 180°; field of view, 380x380mm; acquisition matrix size, 72x72; slice thickness, 3.5mm; no interslice gap; ASSET (SENSE) parallel imaging factor, 3; 3 phase offsets; and a 6.45-ms, 3.2 G/cm, bipolar, trapezoidal, motion-encoding gradient (MEG) on each side of the refocusing pulse.

To obtain a consistent position of the liver for each phase offset, patients were instructed to hold their breath at the end of expiration. A stiffness map (elastogram) for each MR elastography slice was automatically generated by processing the acquired images of propagating shear waves using a previously described 3D local frequency estimation (LFE) inversion algorithm [1]. Elastograms were generated both in gray scale and as a color bit-map with a scale corresponding to the stiffness values. The MEGs were applied sequentially in 3 orthogonal directions to measure the vector displacement field throughout the volume. 3D/3-axis MR elastography provides theoretical advantages of enhanced accuracy of elastograms over 2D MR elastography, as it allows for the removal of longitudinal wave propagation with curl filtering and 3D processing reduces artifacts from through-plane wave propagation. Multislice acquisitions can sometimes have small interslice phase shifts, independent of the desired phase due to the tissue vibrations that if left in the data can adversely affect 3D processing because these phase variations do not reflect the actual tissue motion. This artifact appears as a high-frequency striping pattern in the slice direction due to the interleaved slice acquisition. To remove this artifact, a 1D lowpass filter was applied in the slice direction of each complex-valued image volume before calculating the curl of the wave field and performing the LFE. The filter was designed as a 4th-order Butterworth lowpass filter with a cut-off frequency of 0.357 cm^-1. Spatial presaturation bands were applied on each side of the imaging volume to reduce motion artifacts derived from blood ﬂow. To prevent chemical shift artifacts, spatial spectral pulses were used to generate the initial 90° radio frequency (RF) pulse. Like in other liver and spleen MR elastography studies, susceptibility artifacts normally associated with EPI-based sequences were not a problem in our study [2].

REFERENCES

1 Manduca A, Oliphant TE, Dresner MA et al (2001) Magnetic resonance elastography: non-invasive mapping of tissue elasticity. Med Image Anal 5:237-254

2 Nedredal GI, Yin M, McKenzie T et al (2011) Portal hypertension correlates with splenic stiffness as measured with MR elastography. J Magn Reson Imaging 34:79-87