With the rapid advancement of artificial intelligence technology, chatbots have shown great potential in the healthcare sector. From personalized health advice to chronic disease management and psychological support, chatbots have demonstrated significant advantages in improving the efficiency and quality of healthcare services. As the scope of their applications expands, the relationship between technological complexity and practical application scenarios has become increasingly intertwined, necessitating a more comprehensive evaluation of both aspects. This paper, from the perspective of he althcare applications, systematically reviews the technological pathways and development of chatbots in the medical field, providing an in-depth analysis of their performance across various medical scenarios. It thoroughly examines the advantages and limitations of chatbots, aiming to offer theoretical support for future research and propose feasible recommendations for the broader adoption of chatbot technologies in healthcare.

ObjectiveTo investigate the differences in abnormal liver biochemical parameters in pregnant patients during the epidemic or non-epidemic period of coronavirus disease 2019 （COVID-19）. MethodsA retrospective analysis was performed for 539 pregnant women who were discharged from Department of Obstetrics， Peking University First Hospital， from October 2017 to March 2022 and had at least one abnormal liver biochemical parameter among alanine aminotransferase （ALT）， aspartate aminotransferase （AST）， gamma-glutamyl transpeptidase （GGT）， total bilirubin （TBil）， and total bile acid. The patients in the epidemic period of COVID-19 and those in the non-epidemic period of COVID-19 were compared in terms of etiology， coagulation parameters， aminotransferases， bile acid， and renal function. The independent samples t-test was used for comparison of continuous data between two groups， and the chi-square test was used for comparison of categorical data between two groups. ResultsAmong the patients discharged from Department of Obstetrics during the non-epidemic period of COVID-19， 262 had abnormal liver biochemical parameters， accounting for 1.46%， while 277 patients had abnormal liver biochemical parameters during the epidemic period of COVID-19， accounting for 1.73% among the patients discharged from Department of Obstetrics during the same period of time， and there was a significant difference between these two groups （χ²=3.947， P=0.047）. The etiological analysis of the patients with abnormal liver biochemical parameters during the two periods showed that there was no difference in the proportion of patients with four pregnancy-specific liver diseases （hyperemesis gravidarum， preeclampsia and eclampsia， intrahepatic cholestasis of pregnancy， and acute fatty liver of pregnancy）. As for the patients with abnormal liver biochemical parameters in pregnancy， there was no significant difference in the proportion of patients with normal creatinine and stimated glomerular filtration rate （eGFR） between the epidemic period and the non-epidemic period （86.78% vs 87.90%， χ²=0.141， P=0.708）. The patients with ALT≥5×upper limit of normal accounted for 7.94% in the epidemic period of COVID-19 and 9.54% in the non-epidemic period （χ²=0.433， P=0.511）， and the patients with severe cholestasis accounted for 7.75% in the epidemic period of COVID-19 and 9.27% in the non-epidemic period （χ²=0.392， P=0.531）. The proportion of patients with obstetric bleeding during the epidemic period of COVID-19 was significantly lower than that during the non-epidemic period （14.61% vs 24.19%， χ²=489.334， P<0.001）. ConclusionThere is no difference in the proportion of patients with pregnancy-specific liver diseases among the patients with abnormal liver biochemical parameters in pregnancy between the epidemic period and the non-epidemic period of COVID-19， and there is no change in the proportion of patients with normal creatinine and eGFR among these patients in the epidemic period of COVID-19.

Objectives:To examine the feasibility of using foundation model in computer vision for echocardiographic left ventricular ejection fraction measurement. Methods:Based on the most extensive publicly accessible repository of echocardiographic loops,EchoNet-Dynamic,featuring 10024 recordings from individual patients,a foundation model in computer vision,VideoMAE V2,was fine-tuned,validated,tested using 7460,1288,and 1276 echocardiographic loops,respectively. Results:The mean absolute error between left ventricular ejection fraction measurements of VideoMAE V2 and expert's measurements was 3.94％ (95％CI:3.79％-4.11％).The Pearson's correlation coefficient was 0.91 (95％CI:0.89-0.92).Additionally,VideoMAE V2 demonstrated exceptional accuracy in identifying patients with a left ventricular ejection fraction below 50％,achieving an AUC of 0.96 (95％CI:0.95-0.97). Conclusions:This study validates the feasibility of using foundation model in computer vision for measuring left ventricular ejection fraction in echocardiographic loops and lays the foundation for the development of a generalized multimodal automated interpretation system for echocardiography.

Objective A knowledge graph construction pipeline of traditional Chinese medicine(TCM)diagnosis and treatment was designed and applied,aiming at the automatic construction of the"disease-symptom-pathogenesis-and medicine"knowledge graph based on the medical records of famous TCM physicians,to analyze,organize and present medical records efficiently.Methods Firstly,The entity extraction method of medical records combining Deep Learning and Regular Expression was designed to extract disease,symptom,pathogenesis,and TCM entities from unstructured medical records automatically;secondly,entity relationships were defined and the correlations between entities were calculated using HAN method,and then the"entity-relation-entity"triplets were built;the graph database Neo4j and Gephi were used for knowledge storage and visual display separately;Finally,the application was verified in the Medical records of lung cancer treated by the old famous TCM physicians.Results The precision,Recall and F1 of the knowledge extraction model for medical records entities extraction are 88.49%,90.02%and 89.25%,respectively,and each index is better than the comparison methods.A total of 1077 triples are extracted through entity correlation calculation,and the knowledge graph is successfully constructed.It can reflect the relationship between 'disease-symptom-pathogenesis-medicine' in the treatment of lung cancer by the famous specialists of TCM.Conclusion The method in this paper can effectively solve the extraction,organization and expression of clinical medical records of famous TCM physicians,and realize the automatic construction process from the text of medical records to the knowledge graph.Relevant research ideas and methods proposed in this paper could provide a reference for the construction of the diagnosis and treatment knowledge graph of famous TCM physicians based on medical records.

ObjectiveTo verify and promote the HIV antibody urine test in a demonstration zone for HIV prevention and control in Xinjiang， and to assess the sensitivity and specificity of the test and the willingness of the volunteers， providing a basis for evaluating and determining whether it was suitable for large-scale use in the region. MethodsStratified cluster sampling was used to select the respondents in a place in Xinjiang for questionnaire survey and sample collecting. Questionnaire results were statistically described， and multivariate logistic regression analysis was used to explore the influencing factors of willingness to use HIV-1 urine test. The results of blood ELISA antibody detection method were used as a reference to calculate the specificity， sensitivity and Kappa value of urinary HIV-1 antibody in the total population and different populations. ResultsA total of 2 504 volunteers were recruited for the urine self-testing. The sensitivity of HIV-1 antibody urine test in the total population was 94.73% （54/57）， the specificity was 100% （2 447）， the positive predictive value was 100% （54/54）， and the negative predictive value was 99.88% （2 447/2 450）. The specificity of each population group was 100% when comparing different populations. The test sensitivity was 100% in pregnant women， prostitutes and drug users， 50% （1/2） in MSM subjects， with a Kappa value of 0.796. The test sensitivity was 80% （8/10） in the volunteers， with a Kappa value of 0.878. In the satisfaction survey， 73.1% population was more likely to receive urinary rapid test than the blood fast test， 93.8% of the population was willing to recommend the rapid urinary detection to other people. More than 70% of the population was satisfied with the rapid urinary detection. The multivariate logistic regression analysis showed that age， degree of education， occupation， marital status， payment method of medical expenses， and income level were the influence factors of the purchase of urine fast reagent for the self-testing （P<0.05）. ConclusionUrinary HIV-1 antibody detection method has good reliability and public satisfaction. It is suitable for large-scale use in Xinjiang.

Objective To determine the effect of crocin regulating miR-139-5p/ATF4 axis on myo-cardial injury and energy metabolism in rats with CHF.Methods A total of 84 male SPF SD rats were randomly divided into sham operation group,model group,crocin low-,medium-and high-dose groups,captopril group and crocin+miR-139-5p inhibitor group,with 12 in each group.Car-diac function indexes,myocardial histomathological morphology,apoptotic rate,myocardial injury indexes,heart failure indexes,inflammatory indexes,oxidative stress indexes,myocardial ATP content,SDH activity,and miR-139-5p and ATF4 mRNA expression levels were detected in rats.The targeting relationship between miR-139-5p and ATF4 was verified.Results Compared with the conditions in the model group,crocin treatment at different doses reduced apoptotic rate of cardiomyocytes,decreased MDA content,LVEDD and LVESD values and cTnI,cTnT,CK-MB,NT-proBNP,TNF-α and IL-1β levels,and declined ATF4 mRNA level,and increased LVEF and LVFS values,SOD activity,ATP content,SDH activity and miR-139-5p level(P＜0.05).Com-pared with the crocin high-dose group,the crocin+miR-139-5p inhibitor group had higher apop-totic rate of cardiomyocytes[(22.68±3.25)％vs(11.94±1.38)％,P＜0.05],increased LVEDD and LVESD value,raised MAD content and cTnI,cTnT,CK-MB,NT-proBNP,TNF-α and IL-1βlevel,and elecated ATF4 mRNA level,and decreased LVEF and LVFS value,SOD activity,ATP content,SDH activity and miR-139-5p level(P＜0.05).There was a targeting relationship be-tween miR-139-5p and ATF4.Conclusion Crocin can improve myocardial injury and energy me-tabolism in CHF rats,which may be related to its regulation of miR-139-5p/ATF4 axis.

It has been revealed that hypoxia is dynamic in hypertrophic scars; therefore, we considered that it may have different effects on hypoxia-inducible factor-1α (HIF-1α) and p53 expression. Herein, we aimed to confirm the presence of a teeterboard-like conversion between HIF-1α and p53, which is correlated with scar formation and regression. Thus, we obtained samples of normal skin and hypertrophic scars to identify the differences in HIF-1α and autophagy using immunohistochemistry and transmission electron microscopy. In addition, we used moderate hypoxia in vitro to simulate the proliferative scar, and silenced HIF-1α or p53 gene expression or triggered overexpression to investigate the changes of HIF-1α and p53 expression, autophagy, apoptosis, and cell proliferation under this condition. HIF-1α, p53, and autophagy-related proteins were assayed using western blotting and immunofluorescence, whereas apoptosis was detected using flow cytometry analysis, and cell proliferation was detected using cell counting kit-8 (CCK-8) and 5-bromo-2'-deoxyuridine (BrdU) staining. Furthermore, immunoprecipitation was performed to verify the binding of HIF-1α and p53 to transcription cofactor p300. Our results demonstrated that, in scar tissue, HIF-1α expression increased in parallel with autophagosome formation. Under hypoxia, HIF-1α expression and autophagy were upregulated, whereas p53 expression and apoptosis were downregulated in vitro. HIF-1α knockdown downregulated autophagy, proliferation, and p300-bound HIF-1α, and upregulated p53 expression, apoptosis, and p300-bound p53. Meanwhile, p53 knockdown induced the opposite effects and enhanced HIF-1α, whereas p53 overexpression resulted in the same effects and reduced HIF-1α. Our results suggest a teeterboard-like conversion between HIF-1α and p53, which is linked with scar hyperplasia and regression.
Apoptosis ; Autophagy ; Cell Hypoxia ; Fibroblasts/metabolism* ; Humans ; Hypoxia-Inducible Factor 1, alpha Subunit/metabolism* ; Tumor Suppressor Protein p53/metabolism*

【Objective】 To study the detection performance of HBsAg single-ELISA-reactive samples of blood donors. 【Methods】 Two kinds of ELISA reagents from different manufacturers (named as reagents A and B) were used for HBsAg screening. A total of 276 samples, from January 2017 to May 2021, with HBsAg single-ELISA-reactive results were collected for further nucleic acid detection technology (NAT) and chemiluminescence immunoassay (CLIA) testing, to undergo HBV-DNA and five hepatitis B tests, respectively. The relationship between HBsAg single-ELISA-reactivity, NAT and CLIA was statistically analyzed. 【Results】 Among the 276 HBsAg single-ELISA-reactive samples, 14 were NAT reactive, with the positive rate of 5.07% (14/276). Fisher′s exact test was used to compare the compliance of reagents A and B with NAT reactivity, and the difference was not statistically significant (P<0.05). Among 14 HBsAg+ /NAT+ samples retested by CLIA, 2 were HBsAg reactive(14.29%, 2/14), 13 were anti-HBc reactive (92.86%, 13/14), 9 had the quantitative value of anti-HBs <10 mIU/mL, 5 had the quantitative value of anti-HBs between 10 to 100mIU/ mL. A total of 5 serological patterns were detected, and anti-HBe+ /anti-HBc+ pattern was the dominant. There were 262 cases of HBsAg+ /NAT- samples, but only 1 (0.38%, , 1/262) case was HBsAg reactive by CLIA, 100 were anti-HBc reactive (38.17%, 100/262), 144 (54.96%, 144/262) were anti-HBs reactive, and 1 was HBeAg reactive. A total of 8 serological patterns were detected. 【Conclusion】 Most of HBsAg single-ELISA-reactive results are false, and NAT could effectively reduce the residual risk of transfusion transmitted diseases.

Objective:To assess the predictive values of bioelectrical impedance analysis(BIA)-measured body fat indices to abnormal lipid profiles, and to preliminary propose optimal cut-off values of body fat in children and adolescents.Methods:Children and adolescents, aged 6-16 years, were selected from 30 schools (8 primary schools, 21 middle schools and one 12-year education school) in Dongcheng, Tongzhou, Fangshan and Miyun districts of Beijing by adopting a stratified cluster sampling method from November 2017 to January 2018.Questionnaire survey, body mass index(BMI), body fat mass index (FMI), fat mass percentage (FMP) and four lipid profiles were conducted.Results:A total of 14 309 participants, aged (11.0±3.3) years, were enrolled in the analysis, with 49.9% boys.In boys and girls, the percentile values ( P60- P95) fitted by FMI and FMP with K-median-coefficient of variation(LMS) method were taken as the cutting points, and P75 values were selected as the cut-off points of excessive body fat for their better sensitivity, specificity, predictive value and area under curve (AUC) for identification of abnormal lipid profiles.Boys with FMI above P75 accounted for 28% of the total population, and controlling boys with FMI below P75 could prevent dyslipidemia of 8%-57%.FMI in girl population occupied about 26% of the above, and controlling FMI in girl population below this cut-off point may prevent dyslipidemia from 8%-42%.FMP observed similar results to FMI.Assessed by FMI or FMP with P75 cut-off values, adiposity performed better than BMI for recognizing abnormal lipid profiles in boys (AUC: 52.4%-69.6% vs.50.2%-67.1%, P<0.05) rather than in girls ( P>0.05). In addition, when FMI or FMP beyond P90, the specificity of each abnormal lipid profiles was around 90%. Conclusions:The recommend cut-off points for body fat may be to assess children′s adiposity, and can be applied in preventive activities.

Objective:To investigate the relationships between hyperuricemia and the incidence risk for cardiometabolic abnormity in children.Methods:Data were obtained from School-based Cardiovascular and Bone Health Promotion Program. In 2017, a total of 15 391 children aged 6-16 years in Beijing were selected through stratified cluster sampling at baseline survey. Follow-up investigation was conducted in 2019. Logistic regression model was used to analyze the relationships of uric acid quartiles and change in uric acid levels with incidence risks for cardiometabolic abnormity (hypertension, hyperglycemia and dyslipidemia).Results:A total of 8 807 children (4 376 boys, 4 431 girls) were included in the analysis, the average age of the children was (11.1±3.3) years at baseline survey. The adjusted odds ratios ( ORs) and 95% confidence intervals ( CIs) of incidence risk for hypertension in the third and fourth quartiles of the UA were 1.39 (1.11-1.75) and 1.56 (1.19-1.81), respectively. The ORs and 95% CIs of risk for high LDL-C in the second, third and fourth quartiles were 1.88 (1.16-3.05),1.98 (1.23-3.17) and 2.25 (1.42-3.57). The uric acid level increased by one standard deviation, the risk increased by 17% for hypertension and 27% for high LDL-C. The uric acid level increased by 10 μmol/L, the risk increased by 2.1% for hypertension and 2.9% for high LDL-C. The gender-stratified analysis showed that the similar results. The ORs and 95% CIs were 1.32 (1.09-1.60) and 1.50 (1.05-2.16) for hypertension, 1.90 (1.38-2.60) and 2.96 (1.58-5.52) for high TC, 1.78 (1.26-2.51) and 2.84 (1.60-5.03) for high LDL-C in the groups of newly diagnosed hyperuricemia and persistent hyperuricemia. Conclusions:Higher uric acid level was associated with increased incidence risks for hypertension, abnormal TC and LDL-C. Maintaining optimal uric acid level by children might contribute to the early prevention of cardiovascular diseases.