This paper has constructed a traditional Chinese medicine （TCM） specialized disease dataset platform for mixed hemorrhoids based on a multimodal large model， and the preliminary application has been validated. The platform uses StarRocks to establish a four-level data warehouse system， enabling the aggregation， cleaning， and standardization of multi-source heterogeneous data. Using DeepSeek-R1-Distill-Qwen-7B as the base model， domain fine-tuning is performed through low-rank adaptation （LoRA） technology. Combined with LLaMA-3.3 natural language processing and reasoning chain techniques， the platform enables intelligent parsing and structured extraction of unstructured TCM medical records. It accurately identifies six major categories and 28 subcategories of entities， including symptoms and syndromes， with a fine-tuned model F1 score of 93.8%. The platform has established a high-quality specialized disease dataset containing more than 50,000 medical records and has been applied in a real-world study involving 17,831 patients， preliminarily verifying the efficacy of TCM heritage surgery.

Background As a pillar industry in China, the manufacturing sector has a high incidence of non-fatal occupational injuries. The factors influencing non-fatal occupational injuries in this industry are closely related at various levels, including individual, equipment, environment, and management, making the analysis of these influencing factors complex. Objective To identify influencing factors of non-fatal occupational injuries among manufacturing workers, providing a basis for targeted interventions and surveillance. Methods A total of 2243 frontline workers from cable and shipbuilding enterprises were selected as study subjects to investigate the incidence of non-fatal occupational injuries and collect information at four levels: individual, equipment, management, and environment in past 12 months. Data balancing was performed using resampling, and LASSO regression was used to select factors of non-fatal occupational injuries. The influence degree and type of variables were judged based on the magnitude of the estimated coefficients of each variable, where variables with estimated coefficients > 0 are risk factors, and those <0 are protective factors. The area under the receiver operating characteristic (ROC)curve (AUC) was used to test the performance of the model, with an AUC value > 0.7 indicating good model performance. Results Among the 2243 frontline workers, males accounted for 77.7% (1742 out of 2243), with the main age range being 40-49 years old, representing 29.5% (661 out of 2243), 82.7% of the workers (1854 out of 2243) were married, and 55.6% (1248 out of 2243) had a junior middle school education level. The average monthly income for 51.0% (1144 out of 2243) of the workers was between 5000 and 6999 Chinese Yuan. The incidence of non-fatal occupational injuries among the manufacturing workers was 8.4% (189/2243) in the past 12 months. Among the 22 factors associated with the occurrence of non-fatal occupational injuries (P<0.05), 10 were individual-level factors, including gender, smoking, alcohol consumption, colleague relationships, average exercise duration, job burnout, work fatigue, musculoskeletal disorders, cardiovascular diseases, and neurological and sensory organ diseases; 3 were equipment-level factors, including equipment operability, hazardous workpieces, and safety hazards; 5 were environmental-level factors, including low temperatures, special operations, noise, workspace size, and dirty and disorderly environment; and 4 were management-level factors, including daily working hours, weekly working days, overtime, and pre-job technical training. The AUC value of the LASSO regression model was 0.704 and the final model retained a total of 10 variables. Among them, there were 7 risk factors for non-fatal occupational injuries (coefficient > 0), including safety hazards, musculoskeletal disorders, dangerous workpieces, job burnout, dirty and disorderly environment, smoking, and male gender; and 3 protective factors (coefficient < 0), including pre-job technical training, good colleague relationship, and long working days per week. Conclusion Manufacturing enterprises need to focus on the incidence of non-fatal occupational injuries and conduct targeted interventions for non-fatal occupational injuries by controlling potential safety hazards, providing pre-job technical training, reducing dangerous workpieces, rectifying working environment, and reasonably arranging working hours.

[This corrects the article DOI: 10.1016/j.apsb.2022.03.002.].

OBJECTIVES: To explore the therapeutic mechanism of compound Centella asiatica formula (CCA) for alleviating Schistosoma japonicum (Sj)-induced liver fibrosis in mice. METHODS: The active components and targets of CCA were identified using the TCMSP database with cross-analysis of Sj-related liver fibrosis targets. A "drug-component-target-pathway-disease" network was constructed using Cytoscape 3.9.1. Functional enrichment analysis (GO/KEGG) was performed using DAVID. Molecular docking study was carried out to validate interactions between the core targets and the key compounds. For experimental validation of the results, 36 mice were divided into control group, Sj-infected model group, and CCA-treated groups. In the latter two groups, liver fibrosis was induced via abdominal infection with Sj cercariae for 8 weeks, followed by 8 weeks of daily treatment with CCA decoction or saline. Hepatic pathology of the mice was assessedwith HE and Masson staining, and hepatic expressions of collagen-I and collagen-III were detected using immunohistochemistry; serum IL-6 and TNF-α levels were determined with ELISA. Hepatic expressions of TLR4 and MyD88 proteins were analyzed with Western blotting. RESULTS: We identified a total of 107 bioactive CCA components and 791 targets, including 37 intersection targets linked to Sj-induced fibrosis. The core targets included TNF, TP53, JUN, MMP9, and CXCL8, involving the IL-17 signaling, lipid metabolism, TLR4/MyD88 axis, and cancer pathways. Molecular docking study confirmed strong binding affinity between quercetin (a primary CCA component) and TNF/TP53/JUN/MMP9. In Sj-infected mouse models, CCA treatment significantly attenuated hepatic inflammatory cell infiltration, reduced collagen-I and collagen-III deposition, improved tissue architecture, reduced serum IL-6 and TNF-α levels, and downregulated TLR4 and MyD88 expressions in the liver. CONCLUSIONS CCA mitigates Sj-induced liver fibrosis by targeting TNF, TP53, JUN, and MMP9 to modulate the TLR4/MyD88 pathway, thereby suppressing pro-inflammatory cytokine release, inhibiting hepatic stellate cell activation, reducing collagen deposition, and preventing granuloma formation in the liver.
Animals ; Toll-Like Receptor 4/metabolism* ; Mice ; Myeloid Differentiation Factor 88/metabolism* ; Schistosoma japonicum ; Liver Cirrhosis/parasitology* ; Schistosomiasis japonica ; Signal Transduction ; Molecular Docking Simulation ; Inflammation ; Centella/chemistry* ; Drugs, Chinese Herbal/pharmacology* ; Tumor Necrosis Factor-alpha/metabolism*

Persistent and maladaptive drug-related memories represent a key component in drug addiction. Converging evidence from both preclinical and clinical studies has demonstrated the potential efficacy of the memory reconsolidation updating procedure (MRUP), a non-pharmacological strategy intertwining two distinct memory processes: reconsolidation and extinction-alternatively termed "the memory retrieval-extinction procedure". This procedure presents a promising approach to attenuate, if not erase, entrenched drug memories and prevent relapse. The present review delineates the applications, molecular underpinnings, and operational boundaries of MRUP in the context of various forms of substance dependence. Furthermore, we critically examine the methodological limitations of MRUP, postulating potential refinement to optimize its therapeutic efficacy. In addition, we also look at the potential integration of MRUP and neurostimulation treatments in the domain of substance addiction. Overall, existing studies underscore the significant potential of MRUP, suggesting that interventions predicated on it could herald a promising avenue to enhance clinical outcomes in substance addiction therapy.
Humans ; Substance-Related Disorders/psychology* ; Memory Consolidation/physiology* ; Animals ; Extinction, Psychological/physiology*

Objective : To investigate the mechanism of action of benvitimod (BVM) in the treatment of atopic der- matitis (AD) . Methods : HaCaT cells were stimulated by TNF-α and IFN-γ , and the cells were grouped into NC group , TNF-α/IFN-γ group , TNF-α/IFN-γ BVM group , and TNF-α/IFN-γ BVM ML385 group. The AD model of DNCB-induced Balb/c mice was divided into CON group , DNCB group , DNCB + BVM group , and DNCB + TAC group. The efficacy of BVM and its roles in antioxidant and pyroptosis regulation were evaluated. Results: Compared with the control group , BVM inhibited the inflammatory response of HaCaT cells stimulated by TNF-α and IFN-γ , and ameliorated the skin lesions and inflammation in the DNCB-induced AD mouse model ; at the same time , it significantly increased the expression of nuclearfactorerythroid-2-relatedfactor2 (NRF2)-related anti-oxida- tive stress proteins , and significantly reduced the expression of cellular reactive oxygen species (ROS) levels and pyroptosis proteins. At the same time , the levels of NRF2-related antioxidative stress proteins significantly in- creased , and the levels of ROS and pyroptosis proteins significantly decreased. Conclusion BVM activates the NRF2/ROS/NLRP3 pathway to inhibit pyroptosis , thereby reducing the inflammatory response in AD.

Objective : To analyze the epidemiological characteristics of tuberculosis in Urumqi and assess the im⁃pact of the national health examination policy , as well as COVID⁃19 prevention and control measures , on tuberculo⁃sis incidence in Xinjiang. Methods: The seasonal patterns of pulmonary tuberculosis were examined using the seasonal index method and circular distribution analysis. An Interrupted Time Series (ITS , Interrupted Time Series) model was employed to investigate how the national health examination policy and COVID_19 interventions influenced trends in tuberculosis incidence in Urumqi. Results: From 2013 to 2021 , a total of 19 ,049 cases of tuberculosis were reported in Urumqi. The peak incidence day occurred on May 25th , indicating a consistently high rate throughout the year, particularly during spring and summer months. The results of the interrupted time series model indicated that the initial monthly incidence rate of pulmonary tuberculosis in Urumqi was 5. 05 per 100 ,000. From January 2013 to September 2016 , the incidence rate of pulmonary tuberculosis showed a slow upward trend , and the upward trend was significant (β1 = 0. 039 ,P < 0. 05) . After the implementation of the universal health checkup policy , the incidence rate of pulmonary tuberculosis showed a slow downward trend , and the downward trend was significant(β3 = - 0. 109 ,P < 0. 05) . After the outbreak of the COVID_19 epidemic in January 2020 , the incidence rate of pulmonary tuberculosis in Urumqi first dropped rapidly and then entered a slow upward stage , and the upward trend was significant(β5 = 0. 082 , P < 0. 05) . Conclusion It is recommended that public health authorities prioritize timely identification of potential TB patients during periods with elevated incidence rates observed in spring and summer months. While significant progress has been made through implementing national health examinations policies , attention must be directed towards understanding how ongoing epidemics affect TB patient detection. Thus , enhancing surveillance efforts and intervention strategies are essential in adapting to this new normal. Key words tuberculosis ; seasonal index method ; circular distribution method ; national health examination ; COVID_19 ; interruption time series

Objective:To assess common errors in head and neck CT angiography (CTA) image reports using four types of large language models (LLM), namely GPT-4, DeepSeek, ERNIE Bot and SparkDesk, and to assess the feasibility of using existing LLMs to support quality control of radiology reports in Chinese.Methods:The study was a cross-sectional study. Totally 1 000 head and neck CTA image reports were randomly selected using the simple random sampling method from Xuanwu Hospital, Capital Medical University in 2023, including 500 primary reports and 500 finalized reports. Two radiologists collaboratively identified six types of errors in the reports: description errors, writing errors, left-right confusion errors, diagnostic omissions, logical sequence errors, and other errors. The overall quality of the reports was assessed using a 5-point Likert scale. Subsequently, GPT-4, DeepSeek, ERNIE Bot and SparkDesk models were employed to detect the same six types of errors in the imaging reports and to provide overall scoring. The results from manual review were considered the gold standard for calculating F1 score to evaluate model performance. Intra-class correlation coefficients ( ICC) were used to assess the consistency between manual scores and the overall scores from the four LLMs. Results:In the primary imaging reports, the proportions of manually detected errors were as follows: descriptive errors 2.6% (13/500), writing errors 0.6% (3/500), left-right confusion errors 0, diagnostic omissions 6.4% (32/500), logical sequence errors 5.2% (26/500), and other errors 0. In the finalized imaging reports, the proportions of errors across the six categories were 0.2% (1/500), 0, 0, 0, 0, and 0.2% (1/500), respectively. For error detection in the primary imaging reports, the F1 scores of GPT-4 for the six error types were 0.992, 0.997, 0.997, 0.967, 0.980, and 0.992, respectively. DeepSeek achieved F1 scores of 0.980, 0.955, 0.981, 0.920, 0.995, and 0.960; ERNIE Bot scored 0.982, 0.990, 1.000, 0.956, 0.976, and 0.999; and SparkDesk achieved 0.985, 0.995, 1.000, 0.961, 0.982, and 1.000. In the detection of errors in finalized imaging reports, GPT-4′s F1 scores were 0.994, 0.995, 0.998, 0.973, 0.989, and 0.993; DeepSeek scored 0.968, 0.965, 0.985, 0.971, 0.991, and 0.983; ERNIE Bot achieved 0.996, 0.992, 1.000, 0.983, 0.999, and 0.997; and SparkDesk achieved 0.999, 0.999, 1.000, 1.000, 1.000, and 0.999. The consistency between GPT-4, DeepSeek, and SparkDesk models and human ratings was moderate, with ICC values of 0.514, 0.560, and 0.515 respectively (all P0.001); in contrast, the overall score of ERNIE Bot showed poor consistency with human ratings, with an ICC of 0.221 ( P0.001). Conclusion:LLMs demonstrate high accuracy in detecting errors in head and neck CTA imaging reports. The overall scoring of report quality shows moderate consistency with manual assessments, indicating a certain feasibility for automated quality control in reporting.

Objective To explore the association of diabetes status with the development of tuberculosis (TB) among the community population in Shanghai, and to provide evidence for the formulation of tuberculosis prevention and control strategies. Methods This population-based cohort study was based on Shanghai Suburban Adult Cohort and Biobank (SSACB) in China. The baseline data were acquired by questionnaires, physical examinations and blood biochemistry tests. TB incidence was obtained by matching with TB management information system data. A Cox proportional risk model was established to assess the risk of tuberculosis. Results A total of 36 014 research subjects were included, with an average age of 56.3±11.3 years, of which 14 587 (40.5%) were male. Over 6 years of follow-up, 47 individuals progressed to tuberculosis (incidence rate: 19.8 per 100 000 person-year, 95% CI: 14.6 -26.4). An increased risk of TB was observed in participants with newly diagnosed diabetes compared with those without diabetes (adjusted hazard ratio [aHR], 2.73; 95% CI, 1.19 - 6.28). Conclusion The risk of tuberculosis in newly diagnosed diabetic patients is significantly increased, and strengthening tuberculosis screening for this population should be considered in practical work.

BACKGROUND: Gluconeogenesis is a critical metabolic pathway for maintaining glucose homeostasis, and its dysregulation can lead to glycometabolic disorders. This study aimed to identify hub biomarkers of these disorders to provide a theoretical foundation for enhancing diagnosis and treatment. METHODS: Gene expression profiles from liver tissues of three well-characterized gluconeogenesis mouse models were analyzed to identify commonly differentially expressed genes (DEGs). Weighted gene co-expression network analysis (WGCNA), machine learning techniques, and diagnostic tests on transcriptome data from publicly available datasets of type 2 diabetes mellitus (T2DM) patients were employed to assess the clinical relevance of these DEGs. Subsequently, we identified hub biomarkers associated with gluconeogenesis-related glycometabolic disorders, investigated potential correlations with immune cell types, and validated expression using quantitative polymerase chain reaction in the mouse models. RESULTS: Only a few common DEGs were observed in gluconeogenesis-related glycometabolic disorders across different contributing factors. However, these DEGs were consistently associated with cytokine regulation and oxidative stress (OS). Enrichment analysis highlighted significant alterations in terms related to cytokines and OS. Importantly, osteomodulin ( OMD ), apolipoprotein A4 ( APOA4 ), and insulin like growth factor binding protein 6 ( IGFBP6 ) were identified with potential clinical significance in T2DM patients. These genes demonstrated robust diagnostic performance in T2DM cohorts and were positively correlated with resting dendritic cells. CONCLUSIONS Gluconeogenesis-related glycometabolic disorders exhibit considerable heterogeneity, yet changes in cytokine regulation and OS are universally present. OMD , APOA4 , and IGFBP6  may serve as hub biomarkers for gluconeogenesis-related glycometabolic disorders.
Machine Learning ; Humans ; Computational Biology/methods* ; Biomarkers/metabolism* ; Diabetes Mellitus, Type 2/genetics* ; Animals ; Mice ; Gluconeogenesis/physiology* ; Gene Expression Profiling ; Transcriptome/genetics* ; Gene Regulatory Networks/genetics* ; Clinical Relevance