Ki-67 is a key immunohistochemical marker of tumor proliferative activity in hepatocellular carcinoma (HCC); however, non-invasive assessment of Ki-67 status before surgery remains an unmet clinical need. This study aimed to develop and internally validate multi-sequence magnetic resonance imaging (MRI) radiomics models based exclusively on non-contrast sequences for preoperative prediction of Ki-67 expression. This retrospective study included 91 patients with 97 pathologically confirmed HCC lesions ( high Ki-67, ≥ 10
Liver cirrhosis is a major risk factor for hepatocellular carcinoma (HCC), and accurate risk stratification of cirrhotic nodules remains a clinical challenge. During hepatocarcinogenesis, regenerative and dysplastic nodules undergo heterogeneous alterations in hepatocyte function, transporter expression, vascular architecture, and tissue composition. However, conventional magnetic resonance imaging (MRI) assessment of intranodular signal heterogeneity largely relies on subjective visual interpretation and lacks standardized and reproducible quantitative biomarkers. Gadoxetic acid disodium (Gd-EOB-DTPA)-enhanced MRI provides functional information during the hepatobiliary phase by reflecting hepatocyte-specific uptake mediated by organic anion transporting polypeptides. Impaired transporter expression during malignant transformation is associated with reduced contrast uptake and may contribute to spatially heterogeneous contrast distribution. Quantitative T1 mapping enables voxel-wise measurement of absolute T1 relaxation times, thereby providing an objective and quantitative approach to tissue characterization. However, most existing studies have primarily focused on mean T1 values or global reduction rates, while intranodular spatial heterogeneity remains insufficiently explored. MRI-based heterogeneity analysis may complement conventional mean-value analysis by capturing regional variations within lesions. However, direct evidence on T1 mapping-derived heterogeneity in cirrhotic nodules remains limited, and its incremental value over mean T1 values for reflecting hepatocellular function or early malignant transformation has not been established. Quantitative heterogeneity descriptors, including standard deviation, coefficient of variation, histogram-based parameters, and texture features, can objectively characterize the magnitude and distribution of spatial T1 variation and provide a quantitative framework for assessing intralesional heterogeneity. This review summarizes the biological rationale, technical principles, and emerging evidence regarding Gd-EOB-DTPA-enhanced T1 mapping for the assessment of intranodular heterogeneity. It further discusses its potential role in malignant transformation risk stratification of cirrhotic nodules, current limitations, and future directions for clinical translation.
To evaluate the diagnostic performance of ultrasound-derived fat fraction (UDFF) and auto-point shear wave elastography (Auto-pSWE) in bariatric surgery candidates, identify biological variables independently associated with their values and monitor postoperative changes in metabolic dysfunction-associated steatotic liver disease (MASLD) patients. This prospective study enrolled patients who underwent UDFF and Auto-pSWE examinations within 3 days before bariatric surgery (October 2023-February 2025), with intraoperative biopsy as reference. Diagnostic performance of UDFF and Auto-pSWE (which measured shear wave velocity, SWV) were analyzed by receiver operating characteristic analysis. Multivariable generalized linear models identified variables independently associated with UDFF and SWV. Postoperative follow-ups were conducted at 3 and 6 months in MASLD patients. A total of 237 patients (mean age, 29.42 ± 9.37; 98 men) were included. UDFF showed excellent diagnostic performance for steatosis grades (AUCs: 0.991 for ≥S1, 0.911 for ≥S2, and 0.831 for ≥S3), with steatosis percentage as the strongest independent correlate (partial R2=0.302). SWV exhibited moderate diagnostic performance for significant fibrosis (AUC = 0.752), with fibrosis stages and alanine aminotransferase showing similar independent explanatory power (partial R2=0.067 and 0.063, respectively). In 104 MASLD patients, median UDFF decreased substantially by approximately 80
Diffusion-weighted imaging (DWI) is central to bladder cancer assessment with the Vesical Imaging-Reporting and Data System (VI-RADS). Computed diffusion-weighted imaging (cDWI) reduces scan time but has limited signal-to-noise ratio (SNR); its voxelwise variant, voxelwise-computed diffusion-weighted imaging (vcDWI), may overcome this limitation. We evaluated whether vcDWI provides better image quality and shorter interpretation time while maintaining diagnostic accuracy compared with acquired DWI (aDWI) in VI-RADS assessment. This single-center study included 111 consecutive patients with bladder cancer who underwent magnetic resonance imaging between October 2019 and October 2024. Two radiologists (2 and 5 years’ experience) assessed image quality and objective metrics: tumor SNR (SNRtumor), urine-referenced SNR, contrast-to-noise ratio (CNR), and urine-referenced contrast ratio (CR, the ratio of tumor to urine signal intensity); they performed VI-RADS scoring and recorded interpretation times. Diagnostic performance for muscle invasion (VI-RADS ≥ 4) was evaluated against histopathology. vcDWI showed higher urine-referenced SNR and contrast ratio than all aDWI sequences (all p < 0.001), with CNR comparable to b = 1400 s/mm², whereas lesion-only SNRtumor was lower than at b = 700 and b = 1400 s/mm². It received superior subjective scores for lesion clarity and overall image quality (all p < 0.001). Diagnostic performance was comparable, with the difference in area under the curve (AUC) within a 0.10 non-inferiority margin in an exploratory post hoc analysis (AUCs, vcDWI vs. aDWI: 0.86 vs. 0.85 and 0.90 vs. 0.91; one-sided 95
Early-stage liver fibrosis is reversible; thus, accurate staging of liver fibrosis is of great importance.. Liver biopsy is the gold standard for the diagnosis of liver fibrosis; however, its invasive nature limits its clinical application. Accurate noninvasive staging of liver fibrosis remains challenging. Thus, there is an urgent need for a precise and reliable noninvasive diagnostic approach. A topological deep learning model combining topological data analysis and EfficientViT framework (TopoEfficientViT) was developed to effectively capture global contextual relationships in hepatobiliary phase MR images for assessment of staging liver fibrosis. Our retrospective study included patients with chronic liver diseases from June 2018 to July 2025, which was approved by our Institutional Review Board. Our data from a hospital site (Site A) in our institution were divided into a training set (n = 275) and a validation set (n = 75). The test set (n=65) was composed of data from a separate hospital site (Site B) in the same institution. A total of 415 patients (323 males, mean age 56.52±9.54 years; 92 females, mean age 58.86±9.68 years) were enrolled in the study. In the training set, the AUC and accuracy of TopoEfficientViT for staging liver fibrosis F1-2 vs. F3-4 were 0.942 and 0.906, respectively; for staging liver fibrosis F1-3 vs. F4 were 0.957 and 0.891, respectively. In the validation set, the AUC and accuracy for staging liver fibrosis F1-2 vs. F3-4 were 0.849 and 0.76, respectively; for staging liver fibrosis F1-3 vs. F4 were 0.861and 0.813, respectively. In the test set, the AUC and accuracy for staging liver fibrosis F1-2 vs. F3-4 were 0.829 and 0.769, respectively; for staging liver fibrosis F1-3 vs. F4 were 0.791 and 0.785, respectively. The TopoEfficientViT model had high predictive performance for staging liver fibrosis by integrating multi-scale persistent homology topology with convolutional networks.
