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 To investigate the correlation of monocyte count(MONO)to high density lipopro-tein-cholesterol(HDL-C)ratio(MHR)and systemic immune-inflammation index(SII)with diabetic retinop-athy(DR).Methods Patients with type 2 diabetes mellitus(T2DM)admitted to the hospital from June 2020 to May 2023 were selected as the research objects.According to the presence or absence of DR,the patients were divided into non-retinopathy group(NDR group)and DR Group.The differences in basic information,blood routine,and biochemical indexes between the two groups were analyzed,and the MHR and SII were cal-culated.Multivariate Logistic regression was used to analyze the risk factors for DR.Spearman correlation a-nalysis was used to analyze the correlation between risk factors and DR.The receiver operating characteristic(ROC)curve was used to evaluate the value of MHR and SII in predicting DR in T2DM patients.Results A total of 291 T2DM patients were enrolled,including 135 patients in the NDR group and 156 patients in the DR group.Compared with the NDR group,duration of diabetes was significantly prolonged(P<0.05),glycosy-lated hemoglobin(HbA1c),creatinine,fasting plasma glucose(FPG),total cholesterol(TC),platelet count(PLT),MHR and SII were increased(P<0.05),and high density lipoprotein-cholesterol(HDL-C)was de-creased(P<0.05)in the DR Group.Spearman correlation analysis showed that DR was positively correlated with duration of diabetes,FPG,HbA1c,PLT,MHR and SII(P<0.05),and negatively correlated with HDL-C(P<0.05).Multivariate Logistic regression analysis showed that gender(OR=0.151,95%CI 0.052-0.432,P<0.001),history of heavy drinking(OR=7.199,95%CI 2.845-18.216,P<0.001),duration of di-abetes(OR=1.570,95%CI 1.354-1.821,P<0.001),HbA1c(OR=1.218,95%CI 1.013-1.464,P=0.036),MHR(OR=1.054,95%CI 1.028-1.080,P<0.001)and SII(OR=1.002,95%CI 1.001-1.003,P=0.002)were independent influencing factors for DR patients.ROC curve analysis showed that the area un-der the curve(AUC)of MHR and SII in predicting the development of T2DM to DR was 0.696 and 0.567,re-spectively.The AUC of MHR and SII combined in predicting DR was 0.702.Conclusion MHR and SII are closely related to the incidence of DR,and both have certain predictive value for DR,and the predictive value of the combined of MHR and SII is higher.

Objective:To investigate the association of monocyte-to-high-density lipoprotein cholesterol ratio (MHR) with the severity of white matter hyperintensities (WMHs) and its spatial distribution.Methods:Patients admitted to the Department of Neurology, Jiangning Hospital Affiliated to Nanjing Medical University due to various chronic diseases or physical examinations between January 2023 and December 2024 were included retrospectively. Past medical history, clinical and imaging data were collected. The Fazekas scale was used to assess the severity of WMHs. According to the scoring results of periventricular WMHs (PVWMHs) and deep WMHs (DWMHs), WMHs were divided into no/mild group (0-1 points) and moderate/severe group (2-3 points). Multivariate logistic regression analysis was used to determine independent correlation factors for the severity of WMHs, PVWMHs, and DWMHs. Results:A total of 357 patients were included, aged 65.42±9.95 years, with 198 males (55.5%). There were 193 patients (54.1%) in the no/mild group and 164 (45.9%) in the moderate/severe group. Univariate analysis showed that the proportion of patients with hypertension, diabetes, history of cerebral infarction and cerebral hemorrhage, carotid plaque, and age, serum creatinine, monocyte count and MHR in the moderate/severe group were significantly higher than those in the no/mild group (all P<0.05). Multivariate logistic regression analysis showed a significant positive correlation between MHR and the severity of WMHs (odds ratio 3.138, 95% confidence interval 1.042-9.451; P=0.042). Further analysis showed a significant positive correlation between MHR and PVWMHs (odds ratio 3.384, 95% confidence interval 1.111-10.305; P=0.032), but no independent correlation with DWMHs. In addition, age and hypertension, diabetes, history of cerebral infarction and cerebral hemorrhage were significantly positively correlated with the severity of WMHs, PVWMHs and DWMHs. Conclusion:MHR is correlated with the severity of WMHs, and higher MHR is significantly associated with PVWMHs, but not with DWMHs.

Objective:To investigate the correlation between body mass index (BMI)/high-density lipoprotein cholesterol (HDL-C) ratio and cerebral small vessel disease (CSVD) in middle-aged and elderly people.Methods:Consecutive middle-aged and elderly patients (aged ≥40 years) who were hospitalized for chronic disease examinations in the Department of Neurology, Jiangning Hospital Affiliated to Nanjing Medical University between February 2022 and May 2024 were included prospectively. According to the overall burden score of CSVD, they were divided into CSVD group (≥1) and non-CSVD group (0). According to age, they divided into middle-aged group (40-59 years old) and elderly group (≥60 years old). The demographic characteristics and clinical data were collected. Binary multivariate logistic regression analysis was used to determine the independent correlation between BMI/HDL-C ratio and CSVD. Forest plot was used to analyze the correlation between BMI/HDL-C ratio and CSVD in different age groups. Results:A total of 710 patients were included, with an age of 66.0±10.0 years and 361 were males (50.8%). There were 261 patients (36.8%) in the CSVD group and 449 (63.2%) in the non-CSVD group. The BMI/HDL-C ratio in the CSVD group was significantly higher than that in the non-CSVD group (23.60±7.00 vs. 20.78±6.40; P<0.001). Multivariate logistic regression analysis showed that BMI/HDL-C ratio was an independent risk factor for CSVD in middle-aged and elderly populations (odds ratio 1.046, 95% confidence interval 1.027-1.064; P<0.001). There were 475 patients in the elderly group, of which 198 (41.7%) had CSVD; there were 235 patients in the middle-aged group, of which 63 (26.8%) had CSVD. Forest plot analysis showed that the association between BMI/HDL-C ratio and CSVD still had statistical significance in different age groups, but the effect intensity was higher in the elderly group than in the middle-aged group. Conclusion:The BMI/HDL-C ratio is independently correlated with CSVD in middle-aged and elderly population, particularly significant in the elderly population.

Objective: To establish a prediction model for high-risk cardiovascular disease (CVD) among residents aged 35 to 75 years, so as to provide the basis for improving CVD prevention and control measures. Methods: Permanent residents aged 35 to 75 years were selected from Dongcheng District, Beijing Municipality using the stratified random sampling method from 2018 to 2023. Demographic information, lifestyle, waist circumference and blood biochemical indicators were collected through questionnaire surveys, physical examinations and laboratory tests. Influencing factors for high-risk CVD among residents aged 35 to 75 years were identified using a multivariable logistic regression model, and a prediction model for high-risk CVD was established. The predictive effect was evaluated using the receiver operating characteristic (ROC) curve. Results: A total of 6 968 individuals were surveyed, including 2 821 males (40.49%) and 4 147 females (59.51%), and had a mean age of (59.92±9.33) years. There were 1 155 high-risk CVD population, with a detection rate of 16.58%. Multivariable logistic regression analysis showed that gender, age, smoking, central obesity, systolic blood pressure, fasting blood glucose, triglyceride and low-density lipoprotein cholesterol were influencing factors for high-risk CVD among residents aged 35 to 75 years (all P<0.05). The area under the ROC curve of the established prediction model was 0.849 (95%CI: 0.834-0.863), with a sensitivity of 0.693 and a specificity of 0.863, indicating good discrimination. Conclusion The model constructed by eight factors including demographic characteristics, lifestyle and blood biochemical indicators has good predictive value for high-risk CVD among residents aged 35 to 75 years.

This study aims to explore the role and mechanism of berberine in promoting the expression of aquaporin 4(AQP4) in astrocytes by regulating the expression of miR-383-5p in HepG2 cell-derived exosomes under insulin resistance(IR). The IR-HepG2 cell model was established with 1×10~(-6) mol·L~(-1) insulin. With metformin as the positive control, the safe concentrations of berberine and metformin were screened by cell counting kit-8(CCK-8) and lactate dehydrogenase(LDH) leakage assays, and the effect of berberine on the IR of HepG2 cells was evaluated by glucose consumption. NanoSight was used to measure the particle size and concentration of exosomes secreted by HepG2 cells in each group. HepG2 cell-derived exosomes in each group were incubated with astrocytes for 24 h, and the protein and mRNA levels of AQP4 in HA1800 cells were determined by Western blot and qRT-PCR, respectively. qRT-PCR was performed to determine the expression of miR-383-5p in HepG2 cell-derived exosomes and HA1800 cells after co-incubation. Western blotting was employed to determine the expression levels of miRNAs and proteins associated with exosome production and release in HepG2 cells. The results showed that 10 μmol·L~(-1) berberine and 1 mmol·L~(-1) metformin significantly alleviated the IR of HepG2 cells and reduced the concentration of exosomes in HepG2 cells. The exosomes of HepG2 cells treated with berberine and metformin significantly up-regulated the protein and mRNA levels of AQP4 in HA1800 cells. The mRNA level of miR-383-5p in HepG2 cell exosomes and HA1800 cells co-incubated with berberine and metformin decreased significantly. The intervention with berberine and metformin significantly down-regulated the expression of proteins associated with the production of miRNAs(Dicer, Drosha) as well as the production(Alix, Vps4A) and release(Rab35, VAMP3) of exosomes in IR-HepG2 cells. In conclusion, berberine can promote the expression of AQP4 in astrocytes by inhibiting the production and release of miR-383-5p in HepG2-derived exosomes under IR.
Humans ; MicroRNAs/metabolism* ; Berberine/pharmacology* ; Hep G2 Cells ; Exosomes/genetics* ; Aquaporin 4/metabolism* ; Insulin Resistance ; Astrocytes/drug effects*

This study investigated the mechanism by which Jianpi Bushen Yiqi Decoction promotes acetylcholine receptor(AChR) clustering in myasthenia gravis through the Agrin/low-density lipoprotein receptor-related protein 4(LRP4)/muscle-specific receptor tyrosine kinases(MuSK) signaling pathway. A total of 114 female C57BL/6J mice were divided into the normal group, modeling group, and solvent control group. The normal group and the solvent control group were immunized with phosphate-buffered saline(PBS), while the modeling group was established as an experimental autoimmune myasthenia gravis(EAMG) model using the murine-derived AChR-α subunit R97-116 peptide fragment. After successful modeling, the mice were randomly assigned to the model group, the low-, medium-, and high-dose Jianpi Bushen Yiqi Decoction groups, and the prednisone group. After four weeks of continuous treatment, muscle strength was assessed using Lennon scores and grip strength tests. Immunofluorescence staining was conducted on differentiated C2C12 myotubes incubated with a drug-containing serum to observe the number of AChR clusters. The integrity of AChR on myofilaments in mouse gastrocnemius muscles was further assessed by immunofluorescence staining. Hematoxylin-Eosin(HE)staining was applied to examine pathological changes in the gastrocnemius muscles of EAMG mice treated with Jianpi Bushen Yiqi Decoction. Western blot was utilized to detect the expression of key proteins in the Agrin/LRP4/MuSK signaling pathway in both C2C12 myotubes and mouse gastrocnemius muscles. The results demonstrated that compared to the model group, the prednisone group exhibited a significant decrease in the body weights of mice, whereas no significant differences in the body weights of mice were observed among the low-, medium-, and high-dose Jianpi Bushen Yiqi Decoction groups. All treatment groups showed significantly improved grip strength and Lennon scores. Additionally, the formula promoted AChR clustering on myotubes and enhanced AChR integrity in gastrocnemius myofilaments and reduced inflammatory infiltration between muscle tissue and fibrous hyperplasia. Furthermore, Jianpi Bushen Yiqi Decoction upregulated the protein expression of AChRα1, Agrin, and p-MuSK in C2C12 myotubes and increased the protein expression of AChRα1, Agrin, MuSK, p-MuSK, LRP4, and docking protein 7(Dok-7)in gastrocnemius tissue. In conclusion, Jianpi Bushen Yiqi Decoction may promote AChR clustering by targeting key proteins in the Agrin/LRP4/MuSK signaling pathway, thereby improving neuromuscular junction function and enhancing muscle strength.
Animals ; Agrin/genetics* ; Mice ; Drugs, Chinese Herbal/administration & dosage* ; Signal Transduction/drug effects* ; Receptors, Cholinergic/genetics* ; Female ; Mice, Inbred C57BL ; Receptor Protein-Tyrosine Kinases/genetics* ; Neuromuscular Junction/metabolism* ; Myasthenia Gravis, Autoimmune, Experimental/physiopathology* ; Humans ; LDL-Receptor Related Proteins

The differences in chemical quality characteristics between Xinjiang Rehmannia glutinosa and R. glutinosa were analyzed to provide a theoretical basis for the introduction and quality control of R. glutinosa. In this study, the high performance liquid chromatography(HPLC) fingerprints of 6 batches of Xinjiang R. glutinosa and 10 batches of R. glutinosa samples were established. The content of iridoid glycosides, phenylethanoid glycosides, monosaccharides, oligosaccharides, and polysaccharides in Xinjiang R. glutinosa and R. glutinosa was determined by high performance liquid chromatography-diode array detection(HPLC-DAD), high performance liquid chromatography-evaporative light scattering detection(HPLC-ELSD), and ultraviolet-visible spectroscopy(UV-Vis). The determination results were analyzed with by chemical pattern recognition and entropy weight TOPSIS method. The results showed that there were 19 common peaks in the HPLC fingerprints of the 16 batches of R. glutinosa, and catalpol, aucubin, rehmannioside D, rehmannioside A, hydroxytyrosol, leonuride, salidroside, cistanoside A, and verbascoside were identified. Hierarchical cluster analysis(HCA) and principal component analysis(PCA) showed that Qinyang R. glutinosa, Mengzhou R. glutinosa, and Xinjiang R. glutinosa were grouped into three different categories, and eight common components causing the chemical quality difference between Xinjiang R. glutinosa and R. glutinosa in Mengzhou and Qinyang of Henan province were screened out by orthogonal partial least squares discriminant analysis(OPLS-DA). The results of content determination showed that there were glucose, sucrose, raffinose, stachyose, polysaccharides, and nine glycosides in Xinjiang R. glutinosa and R. glutinosa samples, and the content of catalpol, rehmannioside A, leonuride, cistanoside A, verbascoside, sucrose, and glucose was significantly different between Xinjiang R. glutinosa and R. glutinosa. The analysis with entropy weight TOPSIS method showed that the comprehensive quality of R. glutinosa in Mengzhou and Qinyang of Henan province was better than that of Xinjiang R. glutinosa. In conclusion, the types of main chemical components of R. glutinosa and Xinjiang R. glutinosa were the same, but their content was different. The chemical quality of R. glutinosa was better than Xinjiang R. glutinosa, and other components in R. glutinosa from two producing areas and their effects need further study.
Rehmannia/classification* ; Drugs, Chinese Herbal/chemistry* ; Chromatography, High Pressure Liquid/methods* ; Quality Control

Amid the global wave of digital economy, China's medical artificial intelligence applications are rapidly advancing through technological innovation and policy support, while facing multifaceted evaluation and regulatory challenges. The dynamic algorithm evolution undermines the consistency of assessment criteria, multimodal systems lack unified evaluation metrics, and conflicts persist between data sharing and privacy protection. To address these issues, the China National Health Development Research Center has established a value assessment framework for artificial intelligence medical technologies, formulated the country's first technical guideline for clinical evaluation, and validated their practicality through scenario-based pilot studies. Furthermore, this paper proposes introducing a "regulatory sandbox" model to test technical compliance in controlled environments, thereby balancing innovation incentives with risk governance.
Artificial Intelligence/legislation & jurisprudence* ; China ; Humans ; Algorithms