BackgroundPreoperative ambiguous thyroid nodules often depend on intraoperative frozen sections for surgical planning, but misdiagnosis can occur due to low-quality frozen sections, limited diagnostic time, and a shortage of pathologists. Deep learning models and conventional radiomics have shown potential in improving diagnostic accuracy in thyroid nodules, yet their integration remains under-explored. This study aimed to develop deep-learning-based models to assist in the intraoperative pathological diagnosis of thyroid nodules by classifying benign/malignant cases, predicting BRAFV600E gene mutation, and identifying lymph node metastasis.MethodsA total of 436 Whole-Slide Images (WSIs) of thyroid frozen sections were analyzed using deep learning techniques. The analysis included image preprocessing, feature extraction, and classifier training. Patch-to-WSI feature aggregation was done via Patch Likelihood Histogram (PLH) and Bag of Words (BoW) methods.ResultsOn the test set, the InceptionV3 model performed best in benign/malignant classification with an AUC of 0.998 and accuracy of 0.988, where weakly supervised strategies surpassed supervised ones. For BRAFV600E gene mutation prediction, the ResNet50 model achieved a patch-level AUC of 0.831 and a WSI-level accuracy of 94.4% under the extended strategy. A ViT-based model for lymph node metastasis prediction obtained an AUC of 0.671 and accuracy of 76%.ConclusionsThe study indicates that deep learning models can effectively classify benign/malignant thyroid frozen sections, predict BRAFV600E gene mutations, and predict lymph node metastasis status. It also emphasizes the effectiveness of weakly supervised strategies in thyroid lesion frozen sections, which could lessen reliance on pathologists’ annotations.

BACKGROUND Hepatocellular carcinoma (HCC) is one of the leading causes of cancer-related death worldwide. Dysregulation of the epigenetic modifier PRMT5 contributes to the proliferation and metastasis of cancers, and its inhibition has displayed promising therapeutic effect on malignant tumors. However, the pathogenic mechanism and therapeutic potential of targeting PRMT5 in HCC remain unclear. AIM To elucidate the molecular mechanisms by which PRMT5 promotes HCC progression and to provide a rationale for novel therapeutic strategies targeting PRMT5 . METHODS The correlation between PRMT5 expression and patient prognosis was analyzed using an online database (GEPIA2). Subsequently, the functional impact of PRMT5 was evaluated in vitro through cell viability, clonogenic formation, and apoptosis assays in HCC cell lines. Mechanistically, chromatin immunoprecipitation, co-immunoprecipitation, and promoter activity assays were used to investigate the binding PRMT5 and histone modification (H4R3me2) on the DDIT3 promoter, and its interaction with STAT3 . The anti-tumor efficacy of HLCL-61 was evaluated in subcutaneous xenograft mouse models using both cell line-derived xenografts and patient-derived xenografts. RESULTS Our study found that high PRMT5 expression was correlated with a worse prognosis of HCC, and inhibition of PRMT5 expression significantly decreased the viability of HCC cells by inducing apoptosis. In the mechanistic study, we discovered that PRMT5 could bind to the DDIT3 promoter and increase the H4R3me2 level on it to repress DDIT3 transcription. Meanwhile, PRMT5 interacted with the coiled-coil domain of STAT3 and recruited it to the DDIT3 promoter to conjointly inhibit promoter activity. In addition, we evaluated that the PRMT5 inhibitor HLCL-61 and found that it exhibited excellent inhibitory efficacy on HCC cells and tissue derived tumors. CONCLUSION PRMT5 served as an adaptor of STAT3 , displaying a dual inhibitory role in DDIT3 transcription to promote apoptosis resistance, and its inhibitor HLCL-61 represents a potential alternative therapeutic approach to treat HCC.

ObjectiveThis study investigated how framing effects influence Chinese residents’ willingness to receive the herpes zoster vaccine.MethodsBased on gain-loss framing theory, two versions of the questionnaire were developed: a gain-framed version (Questionnaire A) and a loss-framed version (Questionnaire B). Using convenience sampling, 1,184 participants were enrolled through online and offline channels. Statistical analyses included independent-samples t-tests, one-way analysis of variance (ANOVA), multivariate regression, propensity score matching (PSM), and partial least squares structural equation modeling (PLS-SEM).Results(1) Multivariate regression models showed high explanatory power, with adjusted R2 values of 0.796 (gain-framed), 0.822 (loss-framed), and 0.800 (overall). These values should be interpreted with caution because attitude is conceptually close to vaccination intention. (2) Two chain mediation pathways showed significant indirect associations: framing → information credibility → attitude → vaccination intention (FR → IC → AT→IT) and framing → vaccine price acceptance → attitude → vaccination intention (FR → PC → AT→IT) (both p < 0.001), suggesting that framing was associated with attitude and intention through respondents’ perceived message credibility and price acceptance. (3) A significant framing × gender interaction was observed (B = 0.446, p < 0.001). Stratified regression showed that loss-framed messages were associated with higher vaccination intention among males (β = 0.341; p < 0.001), whereas no statistically significant framing effect was found among females (p = 0.108). (4) A significant framing × herpes zoster (HZ) awareness interaction was also detected (B = 0.313; p = 0.002), indicating that individuals who were aware of HZ responded more positively to loss-framed messages. (5) Monthly income (β = 0.645; p < 0.001) and HZ awareness (β = 0.110; p < 0.001) were identified as important independent predictors.ConclusionFraming effects on herpes zoster vaccination intention differed by gender and disease awareness. Gain-framed messages showed greater explanatory value in the full sample; however, loss-framed messages were associated with stronger vaccination intention among males and HZ-aware individuals, whereas no statistically significant framing effect was observed among females. The chain mediation pathways indicated statistically significant indirect associations among cognition, attitude, and behavioral intention, but did not confirm a psychological mechanism. Monthly income was identified as an important predictor of vaccination intention. These findings may provide a reference for developing more differentiated and targeted vaccine communication strategies.

Objective Kidney stones, affecting approximately 10% of adults worldwide, pose substantial clinical and economic burdens. Although insulin resistance is implicated in lithogenesis, the association between estimated glucose disposal rate, a surrogate marker of insulin sensitivity, and kidney stone risk, as well as the mediating role of lipid markers, remains unexplored. Methods This cross-sectional study analyzed data from 11,282 participants in the National Health and Nutrition Examination Survey (2007–2018). Estimated glucose disposal rate was calculated using waist circumference, hypertension status, and glycosylated hemoglobin. History of kidney stones was self-reported. Weighted logistic regression, restricted cubic spline models, and mediation analysis were used to assess associations, adjusting for sociodemographic, metabolic, and renal covariates. Results In our study, the mean age of participants was 47.71 years, and the prevalence of kidney stones was 9.8%. In the fully adjusted multifactorial logistic regression model, participants in the highest quartile of estimated glucose disposal rate (Q4) were 44% less likely to develop stones than those in Q1 (odds ratio: 0.56, 95% confidence interval: 0.37–0.83; p = 0.005). Restricted cubic spline analysis identified a nonlinear, inverted L-shaped relationship between estimated glucose disposal rate and kidney stone risk, with an inflection point at an estimated glucose disposal rate of 8.72. Below this threshold, each 1-unit increase in estimated glucose disposal rate was associated with an 8% reduction in stone risk (odds ratio: 0.92, 95% confidence interval: 0.88–0.95; p < 0.001), whereas above 8.72, the risk was reduced by 31% for each 1-unit increase (odds ratio: 0.69, 95% confidence interval: 0.60–0.79; p < 0.001). Mediation analysis suggested that high-density lipoprotein mediated 5.72% of the total effect (p = 0.001), independent of body mass index, diabetes, and other confounders. Conclusions Estimated glucose disposal rate and kidney stone risk demonstrated an inverted L-shaped association, with a threshold at 8.72. In addition, high-density lipoprotein partially mediated the association between estimated glucose disposal rate and kidney stones. These findings highlight the interaction between metabolic health and kidney stones and may guide personalized prevention strategies.

BackgroundIn recent years, incidents of public opinion triggered by major public health emergencies have emerged endlessly. Existing studies have focused on public attitudes during the early stages of containment measures but lacked research on how public opinion evolves after those measures are relaxed. In late 2022, however, China optimized its COVID-19 control measures, providing a unique window for this study.ObjectiveTo reveal public attitudes toward the adjustment of response measures for major public health emergencies and how these attitudes evolve over time, and to provide a reference for improving related policies and managing public opinion.MethodsWe collected Baidu Index and Weibo post data related to “epidemic prevention and control” between October 11, 2022 and March 15, 2023. Guided by the “Public Opinion Life Cycle Theory,” we analyzed the evolution of public opinion intensity using the Baidu Index. We applied the SKEP model for sentiment analysis on Weibo posts, exploring changes in public sentiment and differences among groups. Additionally, we used the LDA model for topic mining on Weibo posts, examining the evolution of discussion topics and their underlying causes.ResultsDuring the early stages of adjustments to prevention and control measures, public opinion surged but quickly subsided to a level significantly lower than before, following the announcement of more targeted measures. In the long term, the public generally holds a positive attitude toward these adjustments, though negative sentiment may emerge in the short term. Prior to the adjustments, discussions focused on community prevention and control. In the early phase, debates were intense, with expectations for a return to normal life and economic recovery alongside concerns about health risks and medical resources. After a prolonged adjustment period, discussions on economic and daily-life topics increased, but concerns about medication and reinfection risks remained high.ConclusionTo guide the healthy development of public opinion, policymakers should clearly explain the rationale for policy adjustments, promptly address public concerns, and encourage enterprises and opinion leaders to share positive information; additionally, they should ensure sufficient medical resources are secured before implementing policy changes and roll them out in a well-organized, step-by-step manner.

BACKGROUND:Generative Artificial Intelligence(GenAI) significantly enhances medical research efficiency but raises ethical concerns regarding research integrity. The lack of systematic guidelines for its ethical use underscores the need to investigate GenAI's impact on researchers' awareness and behavior concerning integrity. METHODS/MATERIALS:A cross-sectional survey of 718 valid responses from Chinese medical researchers assessed GenAI's impact on research integrity using an extended Unified Theory of Acceptance and Use of Technology(UTAUT) model. RESULTS:The findings reveal that performance expectancy, effort expectancy, technical environment, trust in technology, and supporting conditions positively influence researchers' awareness of research integrity. Conversely, GenAI anxiety and perceived risks exert a significant negative impact. Furthermore, both supporting conditions and integrity awareness are positively associated with integrity behavior, while GenAI anxiety negatively affects such behavior. CONCLUSION:The stakeholders in the medical research ecosystem should develop comprehensive guidelines for the responsible use of GenAI. Emphasis should be placed on optimizing the technical environment, enhancing trust and support structures, and embedding integrity safeguards, thereby promoting the synergistic development of technological innovation and ethical research practices.

BACKGROUND:The comorbidity of chronic diseases among middle-aged and elderly people is a global public health concern that has attracted great attention in recent years. It is crucial to explore the evolutionary pattern of chronic disease comorbidity in Chinese middle-aged and elderly people and to reveal the developmental trajectory of chronic diseases in this population. METHODS:Data from the China Health and Retirement Longitudinal Study (CHARLS 2015-2020) were utilized for the fixed cohort analysis. Based on the prevalence information of 14 chronic diseases (including hypertension, dyslipidemia, diabetes, cancer, chronic lung diseases, liver disease, heart disease, stroke, kidney disease, stomach diseases, emotional problems, memory-related diseases, arthritis, and asthma) among 10,089 participants aged ≥45 years, association rules and cluster analysis were used to identify trends and trajectories of comorbidities in the middle-aged and elderly populations in China. RESULTS:The analysis revealed that the comorbidity rate of the 14 chronic diseases showed a consistent annual increase from 2015-2020. By 2020, over 85% of patients diagnosed with a single chronic condition exhibited concurrent multimorbidity. This epidemiological progression was paralleled by a progressive increase in detected disease associations: binary comorbidities rose from three significant associations in 2015-10 in 2020, whereas higher-order combinations expanded from one ternary association in 2015-35 ternary and 18 quaternary associations by 2020. Notably, hypertension maintained a central position across all identified comorbidity clusters. The comorbidity patterns identified in 2015 included respiratory, liver and kidney, and cardio-cerebral comorbidity patterns and cancer and emotional problems. The comorbidity patterns identified in 2018 included respiratory, liver and kidney, cerebrovascular, and cardiovascular metabolic comorbidity patterns. The comorbidity pattern in 2020 was the same as that in 2018. CONCLUSION:The issues of comorbidities in chronic diseases among Chinese middle-aged and elderly people is significant, with observed variations in the comorbidity patterns across different time periods. The development of clinical assessment and management guidelines for chronic diseases comorbid with key conditions, such as hypertension and dyslipidemia, is recommended. These guidelines aim to facilitate the co-management, co-treatment, and co-reduction of multiple diseases among middle-aged and elderly people.

IntroductionOnline health communities have become the main source for people to obtain health information. However, the existence of poor-quality health information, misinformation, and rumors in online health communities increases the challenges in governing information quality. It not only affects users’ health decisions but also undermines social stability. It is of great significance to explore the factors that affect users’ ability to discern information in online health communities.MethodsThis study integrated the Stimulus-Organism-Response Theory, Information Ecology Theory and the Mindsponge Theory to constructed a model of factors influencing users’ health information discernment abilities in online health communities. A questionnaire was designed based on the variables in the model, and data was collected. Utilizing Structural Equation Modeling (SEM) in conjunction with fuzzy-set Qualitative Comparative Analysis (fsQCA), the study analyzed the complex causal relationships among stimulus factors, user perception, and the health information discernment abilities.ResultsThe results revealed that the dimensions of information, information environment, information technology, and information people all positively influenced health information discernment abilities. Four distinct configurations were identified as triggers for users’ health information discernment abilities. The core conditions included information source, informational support, technological security, technological facilitation, and perceived risk. It was also observed that information quality and emotional support can act as substitutes for one another, as can informational support and emotional support.DiscussionThis study provides a new perspective to study the influencing factors of health information discernment abilities of online health community users. It can provide experiences and references for online health community information services, information resource construction and the development of users’ health information discernment abilities.

Background: Androgen deprivation therapy (ADT) is the mainstay of treatment for prostate cancer, yet dynamic molecular changes from hormone-sensitive to castration-resistant states in patients treated with ADT remain unclear. Methods: In this study, we combined the dynamic network biomarker (DNB) method and the weighted gene co-expression network analysis (WGCNA) to identify key genes associated with the progression to a castration-resistant state in prostate cancer via the integration of single-cell and bulk RNA sequencing data. Based on the gene expression profiles of CRPC in the GEO dataset, the DNB method was used to clarify the condition of epithelial cells and find out the most significant transition signal DNB modules and genes included. Then, we calculated gene modules associated with the clinical phenotype stage based on the WGCNA. IHC was conducted to validate the expression of the key genes in CRPC and primary PCa patients Results:Nomograms, calibration plots, and ROC curves were applied to evaluate the good prognostic accuracy of the risk prediction model. Results: By combining single-cell RNA sequence data and bulk RNA sequence data, we identified a set of DNBs, whose roles involved in androgen-associated activities indicated the signals of a prostate cancer cell transition from an androgen-dependent state to a castration-resistant state. In addition, a risk prediction model including the risk score of four key genes (SCD, NARS2, ALDH1A1, and NFXL1) and other clinical-pathological characteristics was constructed and verified to be able to reasonably predict the prognosis of patients receiving ADT. Conclusions: In summary, four key genes from DNBs were identified as potential diagnostic markers for patients treated with ADT and a risk score-based nomogram will facilitate precise prognosis prediction and individualized therapeutic interventions of CRPC.

OBJECTIVE:The primary objective of this inquiry was to explore the nexus between authorship attribution in medical literature and accountability for scientific impropriety while assessing the influence of authorial multiplicity on the severity of sanctions imposed. METHODS:Probit regression models were employed to scrutinize the impact of authorship on assuming accountability for scientific misconduct, and unordered multinomial logistic regression models were used to examine the influence of authorship and the number of bylines on the severity of punitive measures. RESULTS:First authors and corresponding authors were significantly more likely to be liable for scientific misconduct than other authors and were more likely to be penalized particularly severely. Furthermore, a negative correlation was observed between the number of authors' affiliations and the severity of punitive measures. CONCLUSION:Authorship exerts a pronounced influence on the attribution of accountability in scientific research misconduct, particularly evident in the heightened risk of severe penalties confronting first and corresponding authors owing to their principal roles. Hence, scientific research institutions and journals must delineate authorship specifications meticulously, ascertain authors' contributions judiciously, bolster initiatives aimed at fostering scientific research integrity, and uphold an environment conducive for robust scientific inquiry.

Background : Acute heart failure (AHF) in the intensive care unit (ICU) is characterized by its criticality, rapid progression, complex and changeable condition, and its pathophysiological process involves the interaction of multiple organs and systems. This makes it difficult to predict in-hospital mortality events comprehensively and accurately. Traditional analysis methods based on statistics and machine learning suffer from insufficient model performance, poor accuracy caused by prior dependence, and difficulty in adequately considering the complex relationships between multiple risk factors. Therefore, the application of deep neural network (DNN) techniques to the specific scenario, predicting mortality events of patients with AHF under intensive care, has become a research frontier. Methods : This research utilized the MIMIC-IV critical care database as the primary data source and employed the synthetic minority over-sampling technique (SMOTE) to balance the dataset. Deep neural network models—backpropagation neural network (BPNN) and recurrent neural network (RNN), which are based on electronic medical record data mining, were employed to investigate the in-hospital death event judgment task of patients with AHF under intensive care. Additionally, multiple single machine learning models and ensemble learning models were constructed for comparative experiments. Moreover, we achieved various optimal performance combinations by modifying the classification threshold of deep neural network models to address the diverse real-world requirements in the ICU. Finally, we conducted an interpretable deep model using SHapley Additive exPlanations (SHAP) to uncover the most influential medical record features for each patient from the aspects of global and local interpretation. Results : In terms of model performance in this scenario, deep neural network models outperform both single machine learning models and ensemble learning models, achieving the highest Accuracy, Precision, Recall, F1 value, and Area under the ROC curve, which can reach 0.949, 0.925, 0.983, 0.953, and 0.987 respectively. SHAP value analysis revealed that the ICU scores (APSIII, OASIS, SOFA) are significantly correlated with the occurrence of in-hospital fatal events. Conclusions : Our study underscores that DNN-based mortality event classifier offers a novel intelligent approach for forecasting and assessing the prognosis of AHF patients in the ICU. Additionally, the ICU scores stand out as the most predictive features, which implies that in the decision-making process of the models, ICU scores can provide the most crucial information, making the greatest positive or negative contribution to influence the incidence of in-hospital mortality among patients with acute heart failure.

BackgroundThe information epidemic emerged along with the COVID-19 pandemic. While controlling the spread of COVID-19, the secondary harm of epidemic rumors to social order cannot be ignored. ObjectiveThe objective of this paper was to understand the characteristics of rumor dissemination before and after the pandemic and the corresponding rumor management and debunking mechanisms. This study aimed to provide a theoretical basis and effective methods for relevant departments to establish a sound mechanism for managing network rumors related to public health emergencies such as COVID-19. MethodsThis study collected data sets of epidemic rumors before and after the relaxation of the epidemic prevention and control measures, focusing on large-scale network rumors. Starting from 3 dimensions of rumor content construction, rumor propagation, and rumor-refuting response, the epidemic rumors were subdivided into 7 categories, namely, involved subjects, communication content, emotional expression, communication channels, communication forms, rumor-refuting subjects, and verification sources. Based on this framework, content coding and statistical analysis of epidemic rumors were carried out. ResultsThe study found that the rumor information was primarily directed at a clear target audience. The main themes of rumor dissemination were related to the public’s immediate interests in the COVID-19 field, with significant differences in emotional expression and mostly negative emotions. Rumors mostly spread through social media interactions, community dissemination, and circle dissemination, with text content as the main form, but they lack factual evidence. The preferences of debunking subjects showed differences, and the frequent occurrence of rumors reflected the unsmooth channels of debunking. The χ2 test of data before and after the pandemic showed that the P value was less than .05, indicating that the difference in rumor content before and after the pandemic had statistical significance. ConclusionsThis study’s results showed that the themes of rumors during the pandemic are closely related to the immediate interests of the public, and the emotions of the public accelerate the spread of these rumors, which are mostly disseminated through social networks. Therefore, to more effectively prevent and control the spread of rumors during the pandemic and to enhance the capability to respond to public health crises, relevant authorities should strengthen communication with the public, conduct emotional risk assessments, and establish a joint mechanism for debunking rumors.

IntroductionIn the response to and prevention and control of the Novel coronavirus pneumonia, the COVID-19 vaccine does not provide lifelong immunity, and it is therefore important to increase the rate of booster shots of the COVID-19 vaccine. In the field of information health science, research has found that information frames have an impact in changing individual attitudes and health behaviors.ObjectiveThis study focuses on the effects of different influencing factors on the public’s willingness to receive the booster shots of the COVID-19 vaccine under two information frameworks.MethodsAn online questionnaire was conducted to explore the effects of demographic characteristics, personal awareness, social relationships, risk disclosure, perceived booster vaccination protection rate, and duration of protection under the assumption of an information framework. T test and one-way analysis were used to testing the effect of variables.Results(1) The persuasion effect under the gain frame is higher than that under the loss frame (B = 0.863 vs. B = 0.746); (2) There was no significant difference in subjects’ intention of booster vaccination in terms of gender, age, income, occupation, educational background and place of residence. Whether family members received booster vaccination was strongly correlated with their intention of vaccination under the loss framework (p = 0.017, M = 4.63, SD = 0.664). (3) The higher the understanding of COVID-19, the higher the degree of compliance with the government’s COVID-19 prevention and control measures, and the higher the willingness to strengthen vaccination; (4) Risk disclosure has a significant impact on people’s willingness to receive COVID-19 booster shots (M = 2.48, under the loss framework; M = 2.44, under the gain framework); (5) Vaccine protection rate and duration of protection have an impact on people’s willingness to vaccinate. Increased willingness to vaccinate when the protection rate of booster vaccine approaches 90% (M = 4.76, under the loss framework; M = 4.68, under the gain framework). When the vaccine protection period is 2 years, people are more willing to receive a booster vaccine; and the willingness to receive a booster shot is stronger under the loss framework (M = 4.60, SD = 0.721, p = 0.879).ConclusionThe impact of the information framework on COVID-19 vaccination intentions is different, and the disclosure of relevant health information should focus on the impact of the information framework and content on the public’s behavior toward strengthening vaccination. Therefore, in the face of public health emergencies, public health departments, healthcare institutions, and other sectors can consider adopting the Gainful Information Framework tool to disseminate health information to achieve better persuasion and promote public health behavior change enhancing public health awareness, and promoting universal vaccination.

为探讨机器学习方法在电子病历领域应用的研究现状、研究热点与前沿,以2000～2022年中国知网数据库和Web of Science核心合集数据库中关于机器学习在电子病历中应用的相关文献为数据来源,运用CiteSpace软件绘制国家/地区、作者、机构、关键词共现以及关键词突现5个方面科学知识图谱进行可视化对比分析,以便了解国内外研究的差异,为该领域的研究和发展提供参考.

Background: In recent years, major sudden public health incidents have occurred frequently, leading to an endless stream of public opinion events. Existing research has mostly focused on the early attitudes of the public towards implementing public health restrictions, with little research attention to the evolution of online public opinion before and after the adjustment of the relaxation of containment measures. However, The end of 2022 saw China optimize its COVID-19 containment measures, providing a special window for this study. Methods: Collect Baidu search index and Weibo post data related to "epidemic prevention and control" from October 11, 2022 to March 15, 2023, and comprehensively use methods such as sentiment analysis, topic mining, and statistical testing to conduct in-depth research on the spatiotemporal evolution characteristics of public opinion intensity, sentiment tendency, and topic, as well as differences among different groups. Findings: At the beginning of the policy adjustment, the intensity of public opinion increased rapidly, and then decreased to a level much lower than that before the adjustment after further optimization of the policy; Public attitudes towards policy adjustment were positive in the long term but showed some negative emotions in the short term; The topics of concern covered many aspects, such as the number and activity of infected people, economic and social development, etc. Interpretation: The relationship between public emotion and the focus topic varies significantly at different stages, in different regions, and among different groups, When governing such public opinion, differentiated and precise response plans should be developed.Funding: This study was supported by the National Social Science Fund of China (Grant No. 20BTQ081) and the Social Science Fund of Hunan Province, China (Grant No.22YBQ014).Declaration of Interest: The authors declare that they have no conflicts of interest associated with the manuscript.

BACKGROUND Electronic medical records have rich clinical information and knowledge value and are an important data source for clinical decision support systems, but the current utilization of electronic medical record data is low, and there are problems of insufficient structure, lack of standardization, and low quality, which affect the development and application of clinical decision support. OBJECTIVE To explore the current research status, research hotspots and frontiers in the field of clinical decision support based on electronic medical records. METHODS In this paper, bibliometric methods were used to obtain relevant literature using the China National Knowledge Infrastructure database and Web of Science core collection database from 2000-2022 as data sources, and electronic medical records and clinical decision support as key search terms. CiteSpace.6.2.R2 software is used to draw a scientific Knowledge graph of author cooperation, national/regional distribution, institutional cooperation, keyword co-occurrence, clustering and time zone map for visual comparative analysis, Python is used to mine and analyze clustering popularity based on the keyword map, so as to understand the differences between domestic and foreign research and provide reference for research and development in this field. RESULTS The field of clinical decision support based on electronic medical record data shows rapid development, with the United States and China as the main research countries and strong cooperation between domestic and foreign institutions and authors. The keywords mainly involve electronic medical records, artificial intelligence, machine learning, clinical pathways, data mining, etc., reflecting the core themes and frontier directions in the field. CONCLUSIONS This paper provides a reference for further promoting the effective use of electronic medical record data and the innovative development of clinical decision-making systems.

Objective To investigate the relationship among information processing, risk/benefit perception and the COVID-19 vaccination intention of OHCs users with the heuristic-systematic model (HSM). Methods This study conducted a cross-sectional questionnaire via an online survey among Chinese adults. A structural equation model (SEM) was used to examine the research hypotheses. Results Systematic information processing positively influenced benefit perception, and heuristic information processing positively influenced risk perception. Benefit perception had a significant positive effect on users' vaccination intention. Risk perception had a negative impact on vaccination intention. Findings revealed that differences in information processing methods affect users' perceptions of risk and benefit, which decide their vaccination intention. Conclusion Online health communities can provide more systematic cues and users should process information systematically to increase their perceived benefits, consequently increase their willingness to receive COVID-19 vaccine.

介绍医学院校数据科学与大数据技术专业、大数据管理与应用专业的人才培养现状,从课程设计、师资力量、实践教学、就业前景等方面分析医学院校大数据相关专业人才培养特点,针对医学信息专业群教育提出建议.

Oral squamous cell carcinoma (OSCC) is a highly invasive type of head and neck cancer. Circular RNA (circRNA) acts as a competing endogenous RNA (ceRNA) and involves in pathogenesis of many diseases. However, the circRNA-miRNA-mRNA network and relationship between ceRNA and immune infiltration in OSCC remain unknown. In this study, we established a ceRNA network, including 89 circRNAs, 43 miRNAs and 223 mRNAs, and found that 233 genes are mainly related to malignant signalling pathways (including "Integrin family cell surface interactions" and "Epithelial-to-mesenchymal transition" pathways) and five potential biomarkers (SLC20A1, PITX2, hsa-mir-135b, hsa-mir-377 and hsa-let-7c). Meanwhile, we established a prognostic model based on clinical risk, and revealed the relationship between immune infiltrating cells and biomarkers in OSCC. Taken together, our study is helpful to reveal the pathogenesis of oral squamous cell carcinoma.