This prospective, single-center study, approved by the local institutional review board, involved adult patients who were referred to our Hepatology Clinic for liver biopsy and were eligible. All ultrasound elastography (USE) examinations were systematically performed before the liver biopsy procedure as part of the routine pre-biopsy clinical assessment. Similarly, all MRI examinations for VMRE acquisition were completed prior to the liver biopsy, within a standardized interval of no more than 30 days (median: 7 days; range: 1–28 days). At the time of imaging, histopathological results were not yet available. Operators performing USE and readers evaluating VMRE images were blinded to biopsy and clinical data, ensuring that both acquisition and interpretation were conducted without bias. Inclusion criteria included age ≥ 18 years, referral for liver biopsy due to suspected or known chronic liver disease, and eligibility for liver MRI. Exclusion criteria included contraindications to MRI (e.g., pacemakers, metallic implants, claustrophobia), clinical or laboratory evidence of acute hepatitis or cholestasis, hepatic hemosiderosis, and inadequate quality due to artifacts such as low signal-to-noise ratio, motion artifacts, or image distortion.

Liver biopsies were performed under local anesthesia using an ultrasound-guided approach targeting the right hepatic lobe. A semi-automated 18-gauge Tru-Cut biopsy needle was used in all cases. Adequate tissue samples were obtained, with all biopsy specimens containing at least 11 portal tracts to ensure diagnostic reliability. In this study, histopathological evaluation of the presence and degree of hepatic fibrosis and inflammation was accepted as the standard reference. All histological specimens were scored by hepatobiliary pathologists with 16 years of experience. All biopsy samples were stained with hematoxylin-eosin and Masson trichrome. Fibrosis was staged using the METAVIR scoring system (F0-F4), which served as the reference standard for both VMRE and USE. (22). The stages of fibrosis were defined as follows: F0, no fibrosis; F1, portal fibrosis without septa formation; F2, portal fibrosis with a few septa; F3, numerous septa without cirrhosis; F4, cirrhosis. Based on this, patients were categorized into significant and non-significant fibrosis (F0-1 vs. F2-4) and early-stage vs. advanced fibrosis (F0-2 vs. F3-4).

All MRI examinations were performed using a 3T scanner (Philips Ingenia®, Best, Netherlands) with a 16-channel body array coil. Diffusion-weighted imaging (DWI) was acquired using a single-shot echo-planar imaging (ss-EPI) sequence with a free-breathing technique. Detailed acquisition parameters are summarized in Table 1.

USE was performed by a radiology resident with 5 years of experience using the 2D-SWE technique on the Mindray DC-40 (Mindray, Shenzhen, China) device. Measurements were taken from the right lobe in a mild right lateral decubitus position with the right arm maximally abducted, and patients were instructed to maintain a neutral breathing state. USE images were generated on a gray-scale ultrasound background.

To assess intraobserver reproducibility, two independent acquisitions were performed consecutively for each patient by the same radiology resident. In each acquisition, two circular ROIs (10 mm diameter) were manually placed by a single radiologist with 5 years of experience, within the same elasticity color box on an axial image, avoiding large vessels and bile ducts, and ensuring a minimum depth of 2 cm from the liver capsule. Most measurements were obtained at depths between 2 and 5 cm, while in cases such as obesity, measurements up to 6 cm were accepted provided that image quality was adequate. The mean value of the two ROIs from each acquisition was recorded, and the mean of the two acquisitions was used for final analysis. All measurements adhered to an interquartile range (IQR) to median ratio (IQR/median) ≤ 30%, according to international elastography guidelines [23].

The shifted ADC was calculated from DW MRI signals acquired with two optimal b-values of 200 and 1500 s/mm², using the formula:

where sADC is the shifted ADC, S200 is the DW MRI signal with a b-value of 200 s/mm², and S1500​ is the DWI MR signal with a b-value of 1500 s/mm². VMRE value (kPa) was then calculated as:

$$ }\,}\alpha }^}} }\beta $$

The values for α and β (-12.740 and 14.0) reported by Le Bihan et al., which correlate the results with standard MRE, were used from the calibration study reported earlier in a different patient cohort [20]. Diffusion weighted images were processed using FSLeyes software to generate VMRE color map (Fig. 1b) [24]. An abdominal radiologist with 15 years of experience and a radiology resident with 5 years of experience independently analyzed the images of 49 patients. Both observers were blinded to histopathological results and clinical data during the image evaluation to avoid interpretation bias. They drew ROIs, excluding artifacts and vascular structures, on at least three consecutive sections of the right liver lobe using ITK-SNAP 4.0 software shown in (Fig. 1a) [25].

VMRE color map a, VMRE segmentation b

Statistical analyses were performed using SPSS software version 21.0 (Chicago, Illinois, USA). Prior to patient enrollment, an a priori power analysis was conducted using G*Power software. The analysis indicated that a minimum of 40 patients would be required to detect medium effect sizes (Cohen’s d = 0.5) with a power of 80% and a two-sided alpha level of 0.05. Our final sample included 49 patients, exceeding this threshold. The normality of continuous variables was assessed using the Kolmogorov–Smirnov test, with a p-value greater than 0.05 indicating normal distribution. The reliability of the VMRE and USE measurements was evaluated using intraclass correlation coefficients (ICCs), calculated as interobserver agreement for VMRE and intraobserver agreement for USE. Descriptive statistical analyses, including mean, standard deviation (SD), and frequency, were used for parametric data, while the Kruskal-Wallis test was used for non-parametric data, and the one-way ANOVA test was used for parametric comparisons. When the Kruskal-Wallis test revealed statistical significance, Bonferroni-adjusted Dunn post hoc tests were applied to determine which fibrosis stages differed significantly.

In this study, all cases with liver disease of various etiologies and the HBV subgroup were analyzed separately. Receiver operating characteristic (ROC) curve analysis was used to determine the optimal threshold values for VMRE and USE. The optimal cut-off points were determined using the Youden index. Diagnostic sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were calculated. A p-value < 0.05 was considered statistically significant within a 95% confidence interval. Additionally, the diagnostic effectiveness of VMRE was evaluated by comparing the threshold values obtained in this study with previously published standard MRE thresholds and histopathological results [26].