A total of 70 male Sprague-Dawley rats aged 7 to 8 weeks and weighing 260 ± 30g were obtained (China Food and Drug Control Research Institute, Beijing, China ). Animals were housed in the animal facility of the Key Laboratory of Organ Transplantation of Tianjin with unrestricted access to standard chow and water. All experiments were approved by the NANKAI University Animal Experimentation Ethics Committee and performed according to the University's animal care guidelines. Rats were randomly divided into two groups: an MR scan (MR) group (n = 14) and a pathologic analysis (PA) group (n = 56). The MR group was further divided into a 70% PH subgroup (MRph, n = 7) and a control subgroup (MRctrl, n = 7). Before surgery, all 14 rats in the MR group underwent MR imaging, and a random 7 of the 56 rats in the PA group were euthanized for pathologic analysis to establish baseline values. The 7 rats in the MRph subgroup and the 56 rats in the PA group underwent a 70% PH, and the 7 rats in the MRcrtl subgroup underwent a sham operation. MR scan group underwent liver T1 mapping, T2 mapping, and DKI examinations before and on the 1st, 2nd, 3rd, 5th, 7th, 14th, and the 21st day after surgery. The PA group, 56 rats in total, randomly selected seven with 70% PH at each corresponding time point above for pathologic analysis, including Ki-67 indices, hepatocyte diameter, and steatosis grade (Fig. 1).
Anesthesia and surgical procedures
The surgical procedures were performed under inhalation anesthesia. Induction was performed in a glass box with a mixture of 2% isoflurane in oxygen (1.0 L/min) (Ruiwode; Shenzhen, China). During surgery, anesthesia was maintained with 2% isoflurane in oxygen through a mask covering the animal's mouth and nose. The animal placed supine on an operating plate underwent a midline laparotomy, and the liver was freed from its ligaments. For the MRph and PA groups, the common pedicle of the left lateral and median lobes was ligated with a 4-0 suture, and the lobes were resected. The abdomen was then closed with a 3-0 suture. For the MRcrtl group, the left lateral and medial lobes were freed from ligaments. The abdomen was then closed with a 3-0 suture.
Magnetic resonance imaging
All the rats underwent MRI examinations on a 3T MR scanner (MAGNETOM Prisma, Siemens Healthcare, Erlangen, Germany) with an 8-channel animal coil (Chenguang, Shanghai, China). The rats were anesthetized with isoflurane inhalation and placed in a prone position to reduce respiratory motion. T1 weighted imaging (T1WI) was performed for liver segmentation. The T1 mapping was acquired using a 3D gradient echo sequence with the volumetric interpolated breath-hold examination (VIBE) sequence with a dual-flip-angle method. Sequence parameters were: repetition time (TR) = 6.30 ms, echo time (TE) = 2.88 ms, Flip angles = 3° and 15°, field of view (FOV) = 120×98 mm2, slice thickness = 3 mm, matrix =192 × 156, voxel size = 0.6×0.6×3 mm3, bandwidth = 300 Hz/px, acquisition time = 2 min 27sec. T2 mapping was acquired using a multi-echo spin echo sequence, with following parameters: TR = 2000 ms, 6 TEs = 13/26/39/52/65/78 ms, FOV = 120×120 mm2, slice thickness = 3 mm, matrix = 128×96, reconstructed voxel size = 0.5×0.5×3 mm3, bandwidth = 201 Hz/px, acquisition time = 4 min 50 sec. DKI was performed using a free-breathing single-shot EPI sequence with the following parameters: TR = 2300 ms, TE = 74 ms, FOV = 120×98 mm2, slice thickness = 3 mm, matrix = 128×98, reconstructed voxel size = 0.5 × 0.5 × 3 mm3, acquisition time = 4 min 43 sec. Five b-values (0, 500, 1000, 1500, and 2000 s/mm2) were applied in 3 directions.
MRI liver volume measurement
Regions of interest (ROI) were manually drawn using T1WI by two independent blinded examiners. The liver volume was measured using software from an IQQA-3D-Liver workstation (EDDA Technologies, USA), allowing quantitative ROI analysis. Liver volume was determined by summing the manual segmentation from each slice.
Imaging data analysis
Parametric T1 and T2 maps were generated inline after data acquisition. DKI data were post-processed using a prototype software (MR Body Diffusion Toolbox, Siemens Healthcare, Erlangen, Germany), the DKI-derived parameters (D and K values) were obtained. Five ROIs measuring 12~15mm2 were drawn on the residual liver to measure the T1, T2, D, and K values while avoiding large vessels, artifacts, and the liver border. The average of the five values was used as the final ROI. Two experienced radiologists analyzed all images in a blinded manner.
Histology and immunohistochemistry
Liver tissue was fixed in a buffered 4% formaldehyde solution, embedded in paraffin, stained with hematoxylin-eosin (HE) and Sirius red using standard histologic techniques. Liver samples were stained with HE to observe the hepatic pathologic structure and determine the hepatocyte size and steatosis grade. All pathologic specimens were reviewed by an experienced pathologist. Hepatocyte size (diameter) was determined by measuring ten hepatocytes in high-power fields (×400) . The grade of steatosis is based on the proportion of hepatocytes containing visible lipid and is expressed semi-quantitatively on a scale of 0-3 (S0, < 5%; S1, 5%-33%; S2, > 33%-66%; S3, > 66%) . Immunohistochemical staining was performed to determine the proliferative index Ki-67 (1:100; Abcam, Cambridge, UK). The number and ratio of Ki-67-positive cells were determined by manual counting in five high-power random fields (×400) per slide.
Statistical analysis was performed on SPSS Version 22 for Mac (SPSS 22; IBM Company, Chicago, IL). The normal distribution of MR parameters in the MRph and MRctrl groups was tested using the Shapiro-Wilk test. Repeated measure one-way ANOVAs were used to compare MR parameters' difference between two groups at different time points. The liver volume and pathologic indices at different time points were compared by one-way ANOVAs. The correlations between MR and liver regeneration parameters were determined using Pearson correlation coefficient analysis or Spearman's rank correlation analysis. The threshold of significance was set to 0.05.