ObjectiveDiscriminating atypical hepatocellular carcinoma (HCC) from other malignancies in liver nodules classified as Liver Imaging Reporting and Data System category M (LR-M) remains a significant diagnostic challenge on conventional ultrasound examination. The LR-M category, originally intended to capture non-HCC malignancies, paradoxically contains up to 63% of atypical HCCs that deviate from classic enhancement patterns, leading to potential misdiagnosis and suboptimal treatment planning. While deep learning has shown promise in HCC diagnosis, most existing models rely exclusively on single-modality ultrasound, overlooking the diagnostic benefits of integrating complementary information from multiple imaging sources. To address this gap, we propose a novel attention-weighted tri-modal ultrasound network (TUS-Net) that integrates contrast-enhanced ultrasound (CEUS), B-mode ultrasound (BUS), and time-intensity curves (TICs) to improve diagnostic accuracy for these clinically challenging lesions. MethodsOur framework incorporates a three-dimensional convolutional neural network (C3D) backbone to extract spatiotemporal features from CEUS videos, capturing dynamic vascular patterns critical for lesion characterization. To effectively fuse complementary modalities, we introduce a dual-channel feature fusion module (DCFFM) that adaptively combines features from CEUS and BUS through channel-wise attention mechanisms, allowing the model to dynamically weigh the contribution of each modality based on diagnostic relevance. Additionally, we propose a temporal intensity feature fusion module (TIFFM) that leverages quantitative hemodynamic information from TICs to guide the model’s attention toward diagnostically critical temporal phases, such as arterial wash-in and portal venous washout. The model is further enhanced by automated lesion localization using YOLOX and class activation mapping for interpretability, ensuring that predictions align with clinically meaningful imaging features. ResultsEvaluated on a tri-modal ultrasound dataset comprising 161 patients with pathologically confirmed LR-M nodules (131 atypical HCC and 30 non-HCC malignancies), our model achieved an accuracy of 86.83%, a sensitivity of 92.50%, a specificity of 75.50%, and an AUC of 89.32% in screening atypical HCC. Compared to single-modality baselines, TUS-Net demonstrated superior specificity, a clinically critical metric given the higher risk associated with misclassifying non-HCC malignancies. Ablation studies confirmed the contribution of each module, with the full model outperforming both standard C3D and 3D ResNet backbones integrated with attention mechanisms. A reader study involving junior and senior radiologists further validated the clinical utility of AI assistance, showing consistent improvements in specificity and inter-reader consistency, particularly for less experienced clinicians. ConclusionThese results surpass existing benchmark models and demonstrate the potential of our approach to enhance diagnostic precision in clinically specific cases. By intelligently fusing multi-modal ultrasound data with attention-guided mechanisms, TUS-Net offers a reliable and interpretable tool that holds promise for improving the non-invasive diagnosis of atypical HCC in challenging LR-M liver nodules.

ObjectiveDiscriminating atypical hepatocellular carcinoma (HCC) from other malignancies in liver nodules classified as Liver Imaging Reporting and Data System category M (LR-M) remains a significant diagnostic challenge on conventional ultrasound examination. The LR-M category, originally intended to capture non-HCC malignancies, paradoxically contains up to 63% of atypical HCCs that deviate from classic enhancement patterns, leading to potential misdiagnosis and suboptimal treatment planning. While deep learning has shown promise in HCC diagnosis, most existing models rely exclusively on single-modality ultrasound, overlooking the diagnostic benefits of integrating complementary information from multiple imaging sources. To address this gap, we propose a novel attention-weighted tri-modal ultrasound network (TUS-Net) that integrates contrast-enhanced ultrasound (CEUS), B-mode ultrasound (BUS), and time-intensity curves (TICs) to improve diagnostic accuracy for these clinically challenging lesions. MethodsOur framework incorporates a three-dimensional convolutional neural network (C3D) backbone to extract spatiotemporal features from CEUS videos, capturing dynamic vascular patterns critical for lesion characterization. To effectively fuse complementary modalities, we introduce a dual-channel feature fusion module (DCFFM) that adaptively combines features from CEUS and BUS through channel-wise attention mechanisms, allowing the model to dynamically weigh the contribution of each modality based on diagnostic relevance. Additionally, we propose a temporal intensity feature fusion module (TIFFM) that leverages quantitative hemodynamic information from TICs to guide the model’s attention toward diagnostically critical temporal phases, such as arterial wash-in and portal venous washout. The model is further enhanced by automated lesion localization using YOLOX and class activation mapping for interpretability, ensuring that predictions align with clinically meaningful imaging features. ResultsEvaluated on a tri-modal ultrasound dataset comprising 161 patients with pathologically confirmed LR-M nodules (131 atypical HCC and 30 non-HCC malignancies), our model achieved an accuracy of 86.83%, a sensitivity of 92.50%, a specificity of 75.50%, and an AUC of 89.32% in screening atypical HCC. Compared to single-modality baselines, TUS-Net demonstrated superior specificity, a clinically critical metric given the higher risk associated with misclassifying non-HCC malignancies. Ablation studies confirmed the contribution of each module, with the full model outperforming both standard C3D and 3D ResNet backbones integrated with attention mechanisms. A reader study involving junior and senior radiologists further validated the clinical utility of AI assistance, showing consistent improvements in specificity and inter-reader consistency, particularly for less experienced clinicians. ConclusionThese results surpass existing benchmark models and demonstrate the potential of our approach to enhance diagnostic precision in clinically specific cases. By intelligently fusing multi-modal ultrasound data with attention-guided mechanisms, TUS-Net offers a reliable and interpretable tool that holds promise for improving the non-invasive diagnosis of atypical HCC in challenging LR-M liver nodules.

Objective: Accurate evaluation of inflammation severity in ulcerative colitis (UC) can guide treatment strategy selection. The potential value of the pericolic fat attenuation index (FAI) on CT as an indicator of disease severity remains unknown.This study aimed to assess the diagnostic accuracy of pericolic FAI in predicting UC severity. Materials and Methods: This retrospective study enrolled 148 patients (mean age 48 years; 87 males). The fat attenuation on CT was measured in four different locations: the mesocolic vascular side (MS) and opposite side of MS (OMS) around the most severe bowel lesion, the retroperitoneal space (RS), and the subcutaneous area. The fat attenuation indices (FAI MS, FAI OMS, and FAI RS) were calculated as the fat attenuation measured in MS, OMS, and RS, respectively, minus that of the subcutaneous area, and were obtained in the non-enhanced, arterial, and delayed phases. Correlations between the FAI and UC Endoscopic Index of Severity (UCEIS) were assessed using Spearman’s correlation. Predictors of severe UC (UCEIS ≥7) were selected by univariable analysis. The performance of FAI in predicting severe UC was evaluated using the area under the receiver operating characteristic curve (AUC). Results: The FAIMS and FAI OMS scores were significantly higher than FAI RS in three phases (all P < 0.001). The FAIMS and FAI OMS scores moderately correlated with the UCEIS score (r = 0.474–0.649 among the three phases). Additionally, FAI MS and FAI OMS identified severe UC, with AUC varying from 0.77 to 0.85. Conclusion Increased CT attenuation of pericolic adipose tissue could serve as a noninvasive marker for evaluating UC severity. FAI MS and FAI OMS of three phases showed similar prediction accuracies for severe UC identification.

Varicocele, the most common and treatable cause of male infertility, significantly impacts fertility. The pathophysiological mechanisms of varicocele have not been fully understood yet. Recent studies have focused on the pathophysiology of varicocele-induced infertility, highlighting inflammation and oxidative stress as key contributing factors. We reviewed recent research on the roles of inflammation and oxidative stress in the pathophysiology of varicocele and found that they negatively impact semen parameters, spermatogenesis, and testicular and epididymal function. In addition, this article summarizes the related factors of inflammation and oxidative stress caused by varicocele. Finally, a brief consideration on the treatments to address inflammation and oxidative stress is proposed. This review may provide treatment options and targets for varicocele-induced infertility. However, the relationship between inflammation and oxidative stress in varicocele still needs further study.
Varicocele/physiopathology* ; Humans ; Oxidative Stress/physiology* ; Infertility, Male/metabolism* ; Male ; Inflammation/physiopathology* ; Spermatogenesis/physiology*

This study aims to explore the correlation between plasma homocysteine (Hcy) levels and the clinical grading of varicocele (VC) when analyzing the potential pathogenesis of endothelial cells injury by Hcy. A total of 184 VC patients, aged 18-46 years, were included in this study. These patients visited The Second Hospital of Dalian Medical University (Dalian, China), between January 2022 and September 2024. Patients were divided into three groups based on clinical grading: Group A (59 cases, Grade I), Group B (28 cases, Grade II), and Group C (97 cases, Grade III). Additionally, 120 individuals with normal fertility test results during the same period were selected as the control group. Routine blood and biochemical indices were collected from the patients. Differences in clinical indices between groups were compared, and univariate and multivariate linear regression analyses were performed to identify factors associated with clinical grading. The results showed that the median Hcy levels in the control group and in patients with Grade I, II, and III VC were 9.56 (interquartile range [IQR]: 8.66, 14.02) µmol l -1 , 11.28 (IQR: 9.71, 14.55) µmol l -1 , 11.84 (IQR: 10.14, 15.60) µmol l -1 , and 12.27 (IQR: 9.52, 15.40) µmol l -1 , respectively. The differences between the four groups were statistically significant ( χ2 = 12.41, P = 0.006). Multivariate regression analysis indicated that Hcy is a factor associated with the clinical grading of VC ( t = 2.53, P = 0.013). Hcy is associated with the clinical grading and may have clinical value in assessing severity of VC.
Humans ; Varicocele/pathology* ; Male ; Homocysteine/blood* ; Adult ; Middle Aged ; Adolescent ; Young Adult ; Severity of Illness Index ; Case-Control Studies

Objective To investigate the independent risk factors affecting the prognosis of chronic active antibody-mediated rejection (caAMR) after kidney transplantation. Methods A retrospective analysis was conducted on 61 patients who underwent renal biopsy and were diagnosed with caAMR. The patients were divided into caAMR group (n=41) and caAMR+TCMR group (n=20) based on the presence or absence of concurrent acute T cell-mediated rejection (TCMR). The patients were followed up for 3 years. The value of 24-hour urinary protein and estimated glomerular filtration rate (eGFR) at the time of biopsy in predicting graft loss was assessed using receiver operating characteristic (ROC) curves. The independent risk factors affecting caAMR prognosis were analyzed using the LASSO-Cox regression model. The correlation between grouping, outcomes, and Banff scores was compared using Spearman rank correlation matrix analysis. Kaplan-Meier analysis was used to evaluate the renal allograft survival rates of each subgroup. Results The 3-year renal allograft survival rates for the caAMR group and the caAMR+TCMR group were 83% and 79%, respectively. The area under the ROC curve (AUC) for predicting 3-year renal allograft loss was 0.83 ［95% confidence interval (CI) 0.70-0.97］ for eGFR and 0.78 (95% CI 0.61-0.96) for 24-hour urinary protein at the time of biopsy. LASSO-Cox regression analysis and Kaplan-Meier analysis showed that eGFR≤25.23 mL/(min·1.73 m²) and the presence of donor-specific antibody (DSA) against human leukocyte antigen (HLA) class I might be independent risk factors affecting renal allograft prognosis, with hazard ratios of 7.67 (95% CI 2.18-27.02) and 5.13 (95% CI 1.33-19.80), respectively. A strong correlation was found between the Banff chronic lesion indicators of renal interstitial fibrosis and tubular atrophy (P<0.05). Conclusions The presence of HLA class I DSA and eGFR≤25.23 mL/(min·1.73 m²) at the time of biopsy may be independent risk factors affecting the prognosis of caAMR.

ObjectivePrimary liver cancer, predominantly hepatocellular carcinoma (HCC), is a significant global health issue, ranking as the sixth most diagnosed cancer and the third leading cause of cancer-related mortality. Accurate and early diagnosis of HCC is crucial for effective treatment, as HCC and non-HCC malignancies like intrahepatic cholangiocarcinoma (ICC) exhibit different prognoses and treatment responses. Traditional diagnostic methods, including liver biopsy and contrast-enhanced ultrasound (CEUS), face limitations in applicability and objectivity. The primary objective of this study was to develop an advanced, light-weighted classification network capable of distinguishing HCC from other non-HCC malignancies by leveraging the automatic analysis of brightness changes in CEUS images. The ultimate goal was to create a user-friendly and cost-efficient computer-aided diagnostic tool that could assist radiologists in making more accurate and efficient clinical decisions. MethodsThis retrospective study encompassed a total of 161 patients, comprising 131 diagnosed with HCC and 30 with non-HCC malignancies. To achieve accurate tumor detection, the YOLOX network was employed to identify the region of interest (ROI) on both B-mode ultrasound and CEUS images. A custom-developed algorithm was then utilized to extract brightness change curves from the tumor and adjacent liver parenchyma regions within the CEUS images. These curves provided critical data for the subsequent analysis and classification process. To analyze the extracted brightness change curves and classify the malignancies, we developed and compared several models. These included one-dimensional convolutional neural networks (1D-ResNet, 1D-ConvNeXt, and 1D-CNN), as well as traditional machine-learning methods such as support vector machine (SVM), ensemble learning (EL), k-nearest neighbor (KNN), and decision tree (DT). The diagnostic performance of each method in distinguishing HCC from non-HCC malignancies was rigorously evaluated using four key metrics: area under the receiver operating characteristic (AUC), accuracy (ACC), sensitivity (SE), and specificity (SP). ResultsThe evaluation of the machine-learning methods revealed AUC values of 0.70 for SVM, 0.56 for ensemble learning, 0.63 for KNN, and 0.72 for the decision tree. These results indicated moderate to fair performance in classifying the malignancies based on the brightness change curves. In contrast, the deep learning models demonstrated significantly higher AUCs, with 1D-ResNet achieving an AUC of 0.72, 1D-ConvNeXt reaching 0.82, and 1D-CNN obtaining the highest AUC of 0.84. Moreover, under the five-fold cross-validation scheme, the 1D-CNN model outperformed other models in both accuracy and specificity. Specifically, it achieved accuracy improvements of 3.8% to 10.0% and specificity enhancements of 6.6% to 43.3% over competing approaches. The superior performance of the 1D-CNN model highlighted its potential as a powerful tool for accurate classification. ConclusionThe 1D-CNN model proved to be the most effective in differentiating HCC from non-HCC malignancies, surpassing both traditional machine-learning methods and other deep learning models. This study successfully developed a user-friendly and cost-efficient computer-aided diagnostic solution that would significantly enhances radiologists’ diagnostic capabilities. By improving the accuracy and efficiency of clinical decision-making, this tool has the potential to positively impact patient care and outcomes. Future work may focus on further refining the model and exploring its integration with multimodal ultrasound data to maximize its accuracy and applicability.

Objective: Accurate evaluation of inflammation severity in ulcerative colitis (UC) can guide treatment strategy selection. The potential value of the pericolic fat attenuation index (FAI) on CT as an indicator of disease severity remains unknown.This study aimed to assess the diagnostic accuracy of pericolic FAI in predicting UC severity. Materials and Methods: This retrospective study enrolled 148 patients (mean age 48 years; 87 males). The fat attenuation on CT was measured in four different locations: the mesocolic vascular side (MS) and opposite side of MS (OMS) around the most severe bowel lesion, the retroperitoneal space (RS), and the subcutaneous area. The fat attenuation indices (FAI MS, FAI OMS, and FAI RS) were calculated as the fat attenuation measured in MS, OMS, and RS, respectively, minus that of the subcutaneous area, and were obtained in the non-enhanced, arterial, and delayed phases. Correlations between the FAI and UC Endoscopic Index of Severity (UCEIS) were assessed using Spearman’s correlation. Predictors of severe UC (UCEIS ≥7) were selected by univariable analysis. The performance of FAI in predicting severe UC was evaluated using the area under the receiver operating characteristic curve (AUC). Results: The FAIMS and FAI OMS scores were significantly higher than FAI RS in three phases (all P < 0.001). The FAIMS and FAI OMS scores moderately correlated with the UCEIS score (r = 0.474–0.649 among the three phases). Additionally, FAI MS and FAI OMS identified severe UC, with AUC varying from 0.77 to 0.85. Conclusion Increased CT attenuation of pericolic adipose tissue could serve as a noninvasive marker for evaluating UC severity. FAI MS and FAI OMS of three phases showed similar prediction accuracies for severe UC identification.

Objective: Accurate evaluation of inflammation severity in ulcerative colitis (UC) can guide treatment strategy selection. The potential value of the pericolic fat attenuation index (FAI) on CT as an indicator of disease severity remains unknown.This study aimed to assess the diagnostic accuracy of pericolic FAI in predicting UC severity. Materials and Methods: This retrospective study enrolled 148 patients (mean age 48 years; 87 males). The fat attenuation on CT was measured in four different locations: the mesocolic vascular side (MS) and opposite side of MS (OMS) around the most severe bowel lesion, the retroperitoneal space (RS), and the subcutaneous area. The fat attenuation indices (FAI MS, FAI OMS, and FAI RS) were calculated as the fat attenuation measured in MS, OMS, and RS, respectively, minus that of the subcutaneous area, and were obtained in the non-enhanced, arterial, and delayed phases. Correlations between the FAI and UC Endoscopic Index of Severity (UCEIS) were assessed using Spearman’s correlation. Predictors of severe UC (UCEIS ≥7) were selected by univariable analysis. The performance of FAI in predicting severe UC was evaluated using the area under the receiver operating characteristic curve (AUC). Results: The FAIMS and FAI OMS scores were significantly higher than FAI RS in three phases (all P < 0.001). The FAIMS and FAI OMS scores moderately correlated with the UCEIS score (r = 0.474–0.649 among the three phases). Additionally, FAI MS and FAI OMS identified severe UC, with AUC varying from 0.77 to 0.85. Conclusion Increased CT attenuation of pericolic adipose tissue could serve as a noninvasive marker for evaluating UC severity. FAI MS and FAI OMS of three phases showed similar prediction accuracies for severe UC identification.