2.1 Research objects
Thirty male Wistar rats were obtained from Experimental Animal breeding Co., Ltd (China, Jinan). The rats were 7 weeks old and weighed 170-200g. All rats were acclimatized in a Specific Pathogen Free (SPF) environment for one week before the experiment. At 8, 13, 18, 23, 28, and 33 weeks, 5 rats were randomly selected for MRI examination under anesthesia (3ml of Urethane intraperitoneal injection, Shanghai Shanpu Chemical Co., Ltd), then euthanized and sent for histological analysis. A 0.2g/mL solution of Urethane was prepared using saline (total of 10ml/kg) with intraperitoneal injection.The study was approved by the Institutional Animal Care and Use Committee of Shandong First Medical University and was performed in accordance with the National Institutes of Health guidelines for the use of laboratory animals.
2.2 MR Imaging Techniques
All MR imaging was performed using a 3.0-T MR (GE discovery MR750) and utilizing a matched eight-channel animal coil (Wankang Medical Technology Co., Ltd, China). Four standard MR imaging sequences were performed. (A) Axial T2 fat-saturated (FS) images [echo time (TE)/repetition time (TR), 96.1ms/3000ms; echo train length,16]; (B) Coronal T1 FS fast spin-echo (FSE) (TE/TR, 13.5ms/500ms; echo train length, 3); (C) IVIM-DWI: repetition time/echo time (TR/TE) 4000/66.8 ms, slice thickness 3.0mm, matrix 64×64,FOV 140mm×112mm, spatial resolution 3.8mm2, a total of 12 b-values were used: 10, 20, 30, 50, 80, 100, 200, 300, 600, 800, 1000, 1500s/mm2. (D) DCE-MRI: fat-saturated contrast-enhanced T1 images with the liver acquisition with volume acceleration (LAVA) sequence, repetition time/echo time (TR/TE) 5.6/1.9 ms, slice thickness 2.0 mm, matrix 128×128, FOV 200mm×160mm. A total of 80 phases were acquired, with a spatial resolution of 2.0mm2. 1ml of gadopentetate dimeglumine contrast agent (BeiLu Pharmaceutical Co., Ltd, Beijing, China) was administered intravenously (tail vein) at a rate of 0.1ml/s, followed by a 2ml saline (0.9%, Shandong Qidu Pharmaceutical Co., Ltd) flush by hand. The concentration of gadopentetate dimeglumine was 0.5g/ml. After the acquisition of seven baseline dynamic scans, 960 images were collected in total with 80 phases for approximately 5 minutes of scanning; the function tool software of GE MR Advantage Workstation 4.6 was used to perform the measurements of DCE-MRI and IVIM-DWI.
2.3 Image Analysis
2.3.1 Analysis of IVIM Parameters
The DWI signal follows the biexponential model to calculate the signal attenuation IVIM, as :
Where Sb is the signal intensity in the pixel with diffusion gradient b, S0 is the signal intensity without diffusion gradient, D (×10-4 mm2/s) is the water diffusion coefficient in the tissue, f is the perfusion fraction related to microcirculation (flowing blood fraction) and D* (×10-4 mm2/s) is the pseudo-diffusion coefficient which represents perfusion-related diffusion.
2.3.2 Analysis of DCE-MRI Parameters
For DCE-MRI parameter analysis, all data were quantitatively analyzed using image processing software. This was calculated by manually drawing different regions of interest (ROIs), to obtain the time-intensity curves (TICs).
Enhancement factor Fenh (%) = (SImax –SI0) ×100% /SI0;
Enhancement slope Senh (%/s) = (SImax-SI0) 100 / (SI0×Tmax);
where SI0 (baseline signal intensity before contrast injection) approaches the SImax (maximum signal intensity) exponentially in time. SImax is determined as the maximum SI during the DCE-MRI examination and Tmax is the time point of SImax.
2.3.3 Data analysis
The original images were processed using the Advantage Workstation (ADW 4.6 version, GE, US) and post-processed by Functool workstation. Two observers with 15 years and 20 years of experience in MRI were blinded to the information and they individually measured the resulting parameter maps. Any arising disagreements of both radiologists were resolved by consensus. All data were measured 3 times and the average is taken to reduce bias caused by measurement error. On the IVIM-DWI and DCE-MRI series, an ovoid region of interest (ROI) was placed within the bilateral sacral or iliac bone marrow. The ROI of the joint space was placed at the lower third of the cartilaginous portion of the SIJ  and all ROIs were 2-4mm2. The bilateral average values were obtained and care was taken to avoid the cortex, venous plexus, ligaments, or any imaging-related artifacts (Figure 1, 2, 3).
2.4 Histological assessment
After MR examination, the rats were weighed and each rat was euthanized with
1% pentobarbital (Sigma company) 100mg/kg intraperitoneal injection. When the heartbeats were not detected for five minutes, SIJ samples were removed, fixed in 10% formalin (200ml per sample) for one-two days, acid-decalcified with 10% methanoic acid for one week, embedded in paraffin, cut after dehydration in graded ethanol (75% ethanol 15s, 85% ethanol 10s, 95% ethanol 10s, absolute ethanol 1min, absolute ethanol 1min,) and stained with hematoxylin (8min) and eosin (2min). The histological changes of the SIJ were observed under a microscope (MODEL BX53F, OLYMPUS, Tokyo, Japan).
2.5 Statistical Analysis
SPSS Statistics version 19.0 was used to perform statistical analysis. The test for homogeneity of variances was performed using Levene’s tests. All the parameter values were compared by one-way ANOVA. The measured parameters were expressed as means ± standard deviation (SD). P values <0.05 were considered statistically significant.