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.

Acute kidney injury (AKI) has high morbidity and mortality, but effective clinical drugs and management are lacking. Previous studies have suggested that macrophages play a crucial role in the inflammatory response to AKI and may serve as potential therapeutic targets. Emerging evidence has highlighted the importance of forkhead box protein O1 (FoxO1) in mediating macrophage activation and polarization in various diseases, but the specific mechanisms by which FoxO1 regulates macrophages during AKI remain unclear. The present study aimed to investigate the role of FoxO1 in macrophages in the pathogenesis of AKI. We observed a significant upregulation of FoxO1 in kidney macrophages following ischemia-reperfusion (I/R) injury. Additionally, our findings demonstrated that the administration of FoxO1 inhibitor AS1842856-encapsulated liposome (AS-Lipo), mainly acting on macrophages, effectively mitigated renal injury induced by I/R injury in mice. By generating myeloid-specific FoxO1-knockout mice, we further observed that the deficiency of FoxO1 in myeloid cells protected against I/R injury-induced AKI. Furthermore, our study provided evidence of FoxO1's pivotal role in macrophage chemotaxis, inflammation, and migration. Moreover, the impact of FoxO1 on the regulation of macrophage migration was mediated through RhoA guanine nucleotide exchange factor 1 (ARHGEF1), indicating that ARHGEF1 may serve as a potential intermediary between FoxO1 and the activity of the RhoA pathway. Consequently, our findings propose that FoxO1 plays a crucial role as a mediator and biomarker in the context of AKI. Targeting macrophage FoxO1 pharmacologically could potentially offer a promising therapeutic approach for AKI.

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.

The study investigated the intrinsic changes in material basis of Angelicae Sinensis Radix during wine processing by headspace-gas chromatography-ion mobility spectrometry(HS-GC-IMS), headspace-solid phase microextraction-gas chromatography-mass spectrometry(HS-SPME-GC-MS), and ultra-high performance liquid chromatography-quadrupole-orbitrap mass spectrometry(UPLC-Q-Orbitrap-MS) combined with chemometrics. HS-GC-IMS fingerprints of Angelicae Sinensis Radix before and after wine processing were established to analyze the variation trends of volatile components and characterize volatile small-molecule substances before and after processing. Principal component analysis(PCA) and orthogonal partial least squares-discriminant analysis(OPLS-DA) were employed for differentiation and difference analysis. A total of 89 volatile components in Angelicae Sinensis Radix were identified by HS-GC-IMS, including 14 unsaturated hydrocarbons, 16 aldehydes, 13 ketones, 9 alcohols, 16 esters, 6 organic acids, and 15 other compounds. HS-SPME-GC-MS detected 118 volatile components, comprising 42 unsaturated hydrocarbons, 11 aromatic compounds, 30 alcohols, 8 alkanes, 6 organic acids, 4 ketones, 7 aldehydes, 5 esters, and 5 other volatile compounds. UPLC-Q-Orbitrap-MS identified 76 non-volatile compounds. PCA revealed distinct clusters of raw and wine-processed Angelicae Sinensis Radix samples across the three detection methods. Both PCA and OPLS-DA effectively discriminated between the two groups, and 145 compounds(VIP>1) were identified as critical markers for evaluating processing quality, including 4-methyl-3-penten-2-one, ethyl 2-methylpentanoate, and 2,4-dimethyl-1,3-dioxolane detected by HS-GC-IMS, angelic acid, β-pinene, and germacrene B detected by HS-SPME-GC-MS, and L-tryptophan, licoricone, and angenomalin detected by UPLC-Q-Orbitrap-MS. In conclusion, the integration of the three detection methods with chemometrics elucidates the differences in the chemical material basis between raw and wine-processed Angelicae Sinensis Radix, providing a scientific foundation for understanding the processing mechanisms and clinical applications of wine-processed Angelicae Sinensis Radix.
Wine/analysis* ; Gas Chromatography-Mass Spectrometry/methods* ; Chromatography, High Pressure Liquid/methods* ; Angelica sinensis/chemistry* ; Solid Phase Microextraction/methods* ; Drugs, Chinese Herbal/isolation & purification* ; Chemometrics ; Volatile Organic Compounds/chemistry* ; Principal Component Analysis ; Ion Mobility Spectrometry/methods*

Researchers commonly use cyclization recombination enzyme/locus of X-over P1 (Cre/loxP) technology-based conditional gene knockouts of model mice to investigate the functional roles of genes of interest in Sertoli and Leydig cells within the testis. However, the shortcomings of these genetic tools include high costs, lengthy experimental periods, and limited accessibility for researchers. Therefore, exploring alternative gene silencing techniques is of great practical value. In this study, we employed adeno-associated virus (AAV) as a vector for gene silencing in Sertoli and Leydig cells. Our findings demonstrated that AAV serotypes 1, 8, and 9 exhibited high infection efficiency in both types of testis cells. Importantly, we discovered that all three AAV serotypes exhibited exquisite specificity in targeting Sertoli cells via tubular injection while demonstrating remarkable selectivity in targeting Leydig cells via interstitial injection. We achieved cell-specific knockouts of the steroidogenic acute regulatory ( Star ) and luteinizing hormone/human chorionic gonadotropin receptor (Lhcgr) genes in Leydig cells, but not in Sertoli cells, using AAV9-single guide RNA (sgRNA)-mediated gene editing in Rosa26-LSL-Cas9 mice. Knockdown of androgen receptor ( Ar ) gene expression in Sertoli cells of wild-type mice was achieved via tubular injection of AAV9-short hairpin RNA (shRNA)-mediated targeting. Our findings offer technical approaches for investigating gene function in Sertoli and Leydig cells through AAV9-mediated gene silencing.
Animals ; Male ; Leydig Cells/metabolism* ; Mice ; Dependovirus/genetics* ; Sertoli Cells/metabolism* ; Gene Silencing ; Genetic Vectors ; Testis/cytology*

Objective:To develop a modified University of Wisconsin preservation solution (UW solution) containing α 2-adrenergic receptor agonists (dexmedetomidine) and noble gases (argon) and investigate its protective effect on the isolated amputated skeletal muscle of rats. Methods:Sixty male SD rats were selected to establish a hindlimb cold preservation/perfusion model and were divided into blank control group, hypothermic storage group, UW solution perfusion group, and modified UW solution perfusion group using a random number table, with 15 rats in each group. Simultaneously, a cold preservation model of rat skeletal muscle myoblasts (L6 cells) was established and the rats were also divided into four groups in the same way. Animal models were prepared in different ways: In the blank control group, the hindlimbs received no special treatment; In the hypothermic storage group, the amputated hindlimbs were stored in a dry centrifuge tube at 4℃ for 18 hours; In the UW solution perfusion group, the amputated hindlimbs were perfused with UW solution and then stored in a centrifuge tube containing UW solution at 4℃ for 18 hours; In the modified UW solution perfusion group, the amputated hindlimbs were perfused with modified UW solution (containing 0.1 nmol/L dexmedetomidine and 50% volume fraction of argon) and then stored in a centrifuge tube containing the modified UW solution at 4℃ for 18 hours. Cell models were treated as follows: In the blank control group, L6 cells were cultured under standard conditions; In the hypothermic storage group and UW solution group, L6 cells were treated with conventional culture medium or UW solution, stored in argon-filled sealed bags at 4℃ for 8 hours, and then rewarmed and cultured for 6 hours; In the modified UW solution group, L6 cells were treated with the modified solution, stored in argon-filled sealed bags at 4℃ for 8 hours, and then rewarmed and cultured for 6 hours. After sample collection, skeletal muscle morphology, tissue edema and ultrastructure features were assessed by HE staining, wet-to-dry weight ratio, and transmission electron microscopy, respectively. Additionally, L6 cell morphology was examined by light microscopy. L6 cell viability was determined by cell counting kit-8 (CCK-8) assay (expressed as absorbance A value). Expression levels of glutathione peroxidase 4 (GPX4) protein in both skeletal muscle tissue and L6 cells were evaluated by immunofluorescence staining and Western blot, respectively.Results:After 18 hours of in vitro preservation of rat isolated amputated limbs, the following results were obtained: (1) HE staining results showed that the muscle fiber morphology of the modified UW solution perfusion group was close to that of the blank control group. Moreover, the area ratio of skeletal muscle cells in the modified UW solution perfusion group was significantly higher than that in the hypothermic storage group and UW solution perfusion group ( P<0.05). (2) The wet-dry weight ratio results showed that there was no statistically significant difference among the modified UW solution perfusion group, the blank control group and UW solution group ( P>0.05), with significantly lower ratios in all three groups than that in the hypothermic storage group ( P<0.05). (3) Transmission electron microscopy results revealed that the modified UW solution perfusion group showed no statistically significant differences in ultrastructural metrics, including myofiber diameter, sarcomere length, mitochondrial short-axis/long-axis ratio, and mitochondrial cristae count, compared with those in the blank control group ( P>0.05), and performed significantly better than both the hypothermic storage group and UW solution perfusion group ( P<0.05). (4) Morphological observation of L6 cells showed that the cellular morphology was regular in the modified UW solution perfusion group, close to that in the blank control group, while it was severely damaged in the hypothermic storage group. Moreover, the cells were reduced in number and partially damaged in the UW solution group. The sequence of cell viability expressed as absorbance A value was blank control group >modified UW solution perfusion group > UW solution perfusion group > hypothermic storage group, with statistically significant differences among the four groups ( P<0.05). (5) Immunofluorescence staining showed that there was no statistically significant difference in fluorescence intensity of GPX4 protein expression between the modified UW solution perfusion group and blank control group ( P>0.05), while the fluorescence intensity was higher in the modified UW solution perfusion group than that in the hypothermic storage group and UW solution perfusion group ( P<0.05). Western blot analysis showed that the relative expression level of GPX4 in the modified UW solution group was significantly lower than that in the blank control group ( P<0.05), but higher than that in the hypothermic storage group and UW solution perfusion group ( P<0.05). Conclusion:The modified UW solution can stabilize the expression level of GPX4 protein, thereby inhibiting ferroptosis and alleviating cold preservation injury in both rat amputated isolated limb skeletal muscle tissue and L6 cells.

A centrifugation-free,single-reaction colorimetric method for detection of glucose,utilizing a dual nanozyme cascade system based on gold nanoparticles(AuNPs)and iron-doped carbon dots(FeCDs),was developed in this work.The AuNPs exhibited glucose oxidase-like activity to catalyze glucose oxidation for generation of H2O2,while the FeCDs demonstrated peroxidase-like activity to subsequently catalyze the H2O2-mediated oxidation of 3,3',5,5'-tetramethylbenzidine(TMB).To prevent interference from the blue signal generated by self-aggregation of AuNPs in subsequent quantitative detection,the reaction system was terminated with HCl,converting oxTMB into a stable yellow product.Based on changes in the absorbance at 450 nm of this yellow solution,a quantitative relationship was established between glucose concentration and absorbance at 450 nm(A450).Experimental results demonstrated that this sensor achieved a linear detection range of 44 μmol/L to 11.11 mmol/L(R2=0.993)with a detection limit of 30.68 μmol/L and spiked recoveries of 97.9%-104.7%.By integrating smartphone-based color recognition capabilities,a rapid visual detection platform was established for quantification of glucose through RGB analysis.The validation experimental results using commercial glucose injection samples further confirmed the practical application potential of this methodology.

Objective To understand the changing composition and antibiotic resistance of bacterial species in the clinical isolates from outpatient and emergency department(hereinafter referred to as outpatients)and inpatient children over time in various hospitals,and to provide laboratory evidence for rational antibiotic use.Methods The data on clinically isolated pathogenic bacteria and antimicrobial susceptibility of isolates from outpatients and inpatient children in the CHINET program from 2015 to 2021 were collected and analyzed.Results A total of 278 471 isolates were isolated from pediatric patients in the CHINET program from 2015 to 2021.About 17.1％of the strains were isolated from outpatients,primarily group A β-hemolytic Streptococcus,Escherichia coli,and Staphylococcus aureus.Most of the strains(82.9％)were isolated from inpatients,mainly SS.aureus,E.coli,and H.influenzae.The prevalence of methicillin-resistant S.aureus(MRSA)in outpatients(24.5％)was lower than that in inpatient children(31.5％).The MRSA isolates from outpatients showed lower resistance rates to the antibiotics tested than the strains isolated from inpatient children.The prevalence of vancomycin-resistant Enterococcus faecalis or E.faecium and penicillin-resistant S.pneumoniae was low in either outpatients or inpatient children.S.pneumoniae,β-hemolytic Streptococcus and S.viridans showed high resistance rates to erythromycin.The prevalence of erythromycin-resistant group A β-hemolytic Streptococcus was higher in outpatients than that in inpatient children.The prevalence of β-lactamase-producing H.influenzae showed an overall upward trend in children,but lower in outpatients(45.1％)than in inpatient children(59.4％).The prevalence of carbapenem-resistant Klebsiella pneumoniae(CRKpn),carbapenem-resistant Pseudomonas aeruginosa(CRPae)and carbapenem-resistant Acinetobacter baumannii(CRAba)was 14％,11.7％,47.8％in outpatients,but 24.2％,20.6％,and 52.8％in inpatient children,respectively.The prevalence of multidrug-resistant E.coli,K.pneumoniae,Proteus mirabilis,P.aeruginosa and A.baumannii strains was lower in outpatients than in inpatient children.The prevalence of fluoroquinolone-resistant E.coli,ESBLs-producing K.pneumoniae,ESBLs-producing P.mirabilis,carbapenem-resistant E.coli(CREco),CRKpn,and CRPae was lower in children in outpatients than in inpatient children,but the prevalence of CRAba in 2021 was higher than in inpatient children.Conclusions The distribution of clinical isolates from children is different between outpatients and inpatients.The prevalence of MRSA,ESBL,and CRO was higher in inpatient children than in outpatients.Antibiotics should be used rationally in clinical practice based on etiological diagnosis and antimicrobial susceptibility test results.Ongoing antimicrobial resistance surveillance and prevention and control of hospital infections are crucial to curbing bacterial resistance.