Animal model
This study was approved by the local Animal Experimentation Ethics Committee and performed in accordance with institutional guidelines (No: 2017N030KY). Twenty adult New Zealand White rabbits weighing 2.5-3.0 kg were randomly divided into two groups: hepatic WIRI model group (n = 10) and sham-operation control group (n = 10). All of rabbits were anesthetized with intravenous injection of 5% amobarbital (Sigma-Aldrich Chemical, St. Louis, MO, USA) at a dose of 1 mL/kg. Hepatic WIRI was induced by temporarily clamping the hepatic artery and portal vein for 30 min and reperfusion for 6 h, followed by MRI studies. The control group only underwent laparotomy and liver ligament dissection. All experimental procedures were performed by a well-experienced surgeon (with 15 years of experience in general surgery).
MR imaging
MR data acquisition was performed under 6-channel phased array body coil of 3 T clinical system (MAGNETOM Trio Tim, Siemens Healthcare, Erlangen, Germany). All rabbits were anesthetized with intramuscular injection xylazine hydrochloride (Huamu Animal, Jilin, China) at a dose of 0.2 mL/kg, and then positioned supinely with a belt to reduce the respiratory motion.
Axial IVIM images were acquired using a free-breathing single-shot echo-planar imaging (ss-EPI) prototype sequence with the following parameters: TR/TE of 1000/57.2 ms, EPI factor of 96, voxel size of 1.9 × 1.9 × 4.0 mm3, field of view (FOV) of 180 × 180 mm2, slice thickness of 4 mm, matrix of 96 × 96, with 3 measurements averaged and an acquisition time of 2 min. DW gradients (i.e. 11 b values of 0, 20, 40, 60, 80, 100, 150, 200, 400, 600, and 800 s/mm2) were applied in three orthogonal directions and were subsequently averaged.
Axial DTI images were acquired using a free-breathing ss-EPI prototype sequence with the following parameters: TR/TE of 3800/86 ms, EPI factor of 88, voxel size of 1.4 × 1.4 × 4.0 mm3, FOV of 180 × 180 mm2, slice thickness of 4 mm, matrix of 128 × 128, with two b values of 0 and 500 s/mm2 on 12 diffusion directions, and an acquisition time of 309 s.
All of animals were exposed to room air (78% N2, 20% O2) to maintain normoxia state. Liver BOLD imaging was performed using a multi-echo gradient echo pulse sequence with free-breathing examination to acquire axial T2*-weighted images. Other imaging parameters included: TR of 75 ms, TE of 2.57-24.25 ms (9 echoes), flip angle of 30°, voxel size of 2.0 × 1.6 × 4.0 mm3, FOV of 300 × 225 mm2, matrix of 192 × 154, slice thickness of 4 mm, number of slices 18, and a total examination time of 76 s.
Image analysis
Image analysis was performed with prototype software supplied by the manufacturer (Siemens Healthcare, Erlangen, Germany). Three IVIM-derived parameters (Dslow: the true diffusion as reflected by pure molecular diffusion, Dfast: the pseudodiffusion coefficient related to perfusion, and PF: the perfusion fraction which represents a fractional volume of microcirculation within a voxel) were calculated by a full bi-exponential fitting of the MR signal intensity images of all b values on a pixel-by-pixel basis, using the equation as previously described [
30].
For quantitative analysis of DTI, the mean diffusitivity (MD) and fractional anisotropy (FA) maps were generated automatically as previously described from all diffusion-weighted images on the postprocessing workstation. The range of FA value is from 0 to 1 [
26].
BOLD-derived T2* (ms) and R2* (1/T2*, sec
−1) were calculated by fitting a monoexponential, voxel by voxel, to the signal intensity-echo time curve [
23,
31]. Color-coded parametric images of R2* were obtained using commercial software (Image J, NIH, Bethesda, MD, USA).
All regions of interest (ROIs) were manually positioned by two radiologists (with 10 and 5 years of experience in abdominal MR imaging) on images with b value of 0 s/mm2, and then were copied to the corresponding Dslow, Dfast, PF, MD, FA, and T2* maps. Six circular ROIs were manually drawn on 3 central slices of the liver to avoid the inclusion of large intrahepatic vessels/bile ducts and the margin. The mean size of all ROIs in the liver is 12 pixels ±2. The mean value of the six ROIs was used for the final analysis.
Biochemical and Histopathological analysis
After MR scanning, 4 ml blood was collected from the auricular vein of each animal. The blood was centrifuged to obtain serum. Serum liver function parameters, including alanine transaminase (ALT), aspartate transaminase (AST), and lactate dehydrogenase (LDH) were measured using automatic biochemistry analyzer (HITACHI TAB-40F2, Japan) with standard procedures. The rabbits were then sacrificed by an overdose of 5% amobarbital by intravenous injection, and 1 g liver tissue were collected for each animal. The oxidant/antioxidant, and inflammatory parameters, including malondialdehyde (MDA), superoxide dismutase (SOD), and myeloperoxidase (MPO) of liver tissue were measured using spectrophotometry method (Jiancheng Bio, Nanjing, China).
Liver specimens were fixed in 4% phosphate-buffered formaldehyde and embedded in paraffin. After that, hematoxylin and eosin (H&E) staining was performed on each section for histological analysis. All HE-stained tissue sections were examined microscopically under an optical microscope (Olympus BX50, Japan). The stages of hepatic inflammation/necrosis (I score) and liver fibrosis (F score) were scored by one histopathologist (with 16 years of liver pathology) using the METAVIR classification system [
32] and blinded to the results of the group and MR.
Statistical analysis
Interobserver reproducibility was assessed by measuring intraclass correlation coefficients (ICC) [
33]. Statistical analysis was performed using independent sample
t test, and Mann-Whitney U test for two group comparisons of the functional MR parameters and biochemical parameters. The Pearson and Spearman correlation coefficient was calculated to assess the correlations between functional MR parameters and biochemical parameters. The data were analyzed using SPSS 17.0 software (SPSS, Chicago, IL, USA),
P < 0.05 was considered statistically significant.