Objective To explore the value of improved ResNet18 lightweight deep learning(DL)models for automatically detecting gouty arthritis(GA)based on ultrasonogram of the first metatarsophalangeal joint(MTP1).Methods A total of 2 401 ultrasonograms obtained from 260 patients with suspected gout who underwent MTP1 ultrasound examination were included and divided into training set(1 910 ultrasonograms from 209 cases)and test set(491 ultrasonograms from 51 cases)at the ratio of 4∶1.GA lesions on ultrasonograms were manually labeled.After preprocessing,ResNet18 lightweight network was used to construct DL models for identifying the ultrasonogram category was normal or abnormal(with any manifestation of GA).Five-fold cross-validation method was adopted to evaluate the efficacy of the DL models constructed with 2,3,4 or 6 residual blocks,i.e.model 1,2,3 and 4,respectively,and the computational cost and the amount of parameters of each model were recorded.The efficacy of the models were verified using test set,and the best DL model was screened.Results The computational cost of model 1,2,3 and 4 was 7 558.27,2 963.73,4 012.33 and 6 093.39 M,respectively,while the amount of parameters was 4.61,4.91,4.91 and 5.28 M,respectively.Model 2 had the least computational cost with parameters only slightly more than model 1.In test set,no significant difference of accuracy nor the area under the curve was found among 4 models(all P＞0.05).The sensitivity of model 2 was higher than that of model 3,while its specificity was lower only than that of model 3(both P＜0.05),hence model 2 was the best DL model.Conclusion Improved ResNet18 lightweight DL models could be used for automatically detecting GA based on ultrasonogram of MTP1,among which model 2 was the best one.

Objective Focusing on the problems of unbalanced nursing education resources and insufficient training,it discusses the path of nursing echelon training and nursing education resources optimization in secondary hospitals.Methods Based on nursing questionnaire,interview data and platform logs from 4 secondary hospitals in Hainan Province,Latent Dirichlet Allocation topic modeling,sentiment analysis and Social Network Analysis were used.Identify the core problems of educational resources and training system and propose optimization schemes.Results Topic modeling revealed that insufficient educational resources and information silos were major obstacles to shared platform development.Similarity analysis result showed high consistency in feedback regarding disconnected training content and single training models.Sentiment analysis result indicated that 50％of negative feedback focused on inadequate evaluation and incentive mechanisms.Social Network Analysis validated the significant role of multi-node collaboration in optimizing resource flows.Conclusion Building shared platforms,implementing differentiated training systems,and developing localized nursing faculty are critical pathways to achieving homogenized nursing education,providing valuable references for optimizing regional nursing education.

Objective:To investigate the factors influencing the delayed onset of intrapartum fever following epidural labor analgesia and their impact on maternal and neonatal outcomes.Methods:Select parturients who experienced intrapartum fever following labor analgesia(T≥38.0℃,age≥18 years,singleton pregnancy,ASA classification Ⅱ)between January 1,2021,and December 31,2023.Group them based on the median time of intra-partum fever onset after labor analgesia:those with onset times less than the median were classified as the ear-ly-onset fever group,and those with onset times greater than the median were classified as the late-onset fever group.Using univariate and multivariate Logistic regression analysis to explore factors influencing the delay in in-trapartum fever onset and the pregnancy outcomes of the mothers and newborns in both groups.Results:A total of 253 parturients were included,and the time range of onset of intrapartum fever following epidural labor analgesi-a was 1.83-28.42 hours,with a median fever onset time of 8.00 hours.There were 126 cases in the early-onset group and 127 in the late-onset group.Multivariate Logistic regression analysis indicated that primiparous women,artificial membrane rupture,and neonatal birth weight were independent risk factors for delayed fever onset(OR＞1,P＜0.05),whereas the administration of oxytocin prior to labor analgesia was found to be a protective factor(OR＜1,P＜0.05).The late-onset group exhibited higher levels of white blood cells(WBC),C-reactive pro-tein,longer hospital stays,higher hospitalization costs,greater diagnosis rates of chorioamnionitis,higher NICU ad-mission rates,as well as a higher incidence of neonatal pneumonia,for newborns compared to the early-onset group(P＜0.05).Conclusions:Primiparous women,artificial membrane rupture,and higher neonatal birth weight may be associated with delayed onset of intrapartum fever,while oxytocin administration prior to labor analgesia may offer some protective benefit.The later the onset of intrapartum fever,the worse the clinical outcomes for both mother and infants.

Pasteurella is a genus of zoonotic pathogens that are widely distributed in both healthy and diseased ani-mals.Reports of human infections caused by wild animal bites are rare.This paper reports the first case of Pasteu-rella multocida infection caused by a snow leopard bite in China.The strains were identified using the DL-96 Ⅱ bac-terial identification system,BRUKER automatic microbial mass spectrometer,and Autof ms 1000 mass spectrome-ter.Antimicrobial susceptibility testing was performed using the DL-96 Ⅱ bacterial assay system.After wound de-bridement,surgical incision and drainage,and combined antimicrobial treatment,the patient recovered and was dis-charged from the hospital.This paper aims to improve clinical health care workers'understanding on this bacterium.

Objective To observe the value of follicular thyroid imaging reporting and data system(F-TIRADS)for differentiating benign and malignant follicular thyroid lesions.Methods Totally 502 patients with follicular thyroid lesions were retrospectively enrolled,including 104 patients with single malignant lesion(malignant group,containing 77 follicular thyroid carcinomas[FTC]and 27 follicular variant of papillary thyroid carcinomas[FVPTC])and 398 patients with 416 benign lesions(benign group,containing 197 follicular thyroid adenomas[FTA]and 219 thyroid adenomatous hyperplasia).Ultrasonic features of lesions were recorded,and F-TIRADS scores were assigned by 1 junior and 1 senior ultrasound physicians.Taken histopathology results as gold standard,receiver operating characteristic curve was drawn,the area under the curve(AUC)was calculated to evaluate the efficacy for differentiating benign and malignant follicular thyroid lesions using F-TIRADS.Results Significant differences of composition,internal echo,boundary,calcification and trabecular structure of lesions were found between groups(all P＜0.001).Taken F-TIRADS score≥ 7 as the optimal cut-off value,the sensitivity,specificity,accuracy,positive predictive value and negative predictive value for differentiating benign and malignant follicular thyroid lesions by the junior physician was 76.92％,77.40％,77.31％,93.06％and 45.98％,while by the senior physician was 78.84％,81.25％,80.76％,93.89％and 51.25％,respectively.The efficacy of the latter was higher than of the former(AUC was 0.827 and 0.859,respectively,P＜0.05).Conclusion F-TIRADS could effectively identifying benign and malignant follicular thyroid lesions.

Objective To explore the value of improved ResNet18 lightweight deep learning(DL)models for automatically detecting gouty arthritis(GA)based on ultrasonogram of the first metatarsophalangeal joint(MTP1).Methods A total of 2 401 ultrasonograms obtained from 260 patients with suspected gout who underwent MTP1 ultrasound examination were included and divided into training set(1 910 ultrasonograms from 209 cases)and test set(491 ultrasonograms from 51 cases)at the ratio of 4∶1.GA lesions on ultrasonograms were manually labeled.After preprocessing,ResNet18 lightweight network was used to construct DL models for identifying the ultrasonogram category was normal or abnormal(with any manifestation of GA).Five-fold cross-validation method was adopted to evaluate the efficacy of the DL models constructed with 2,3,4 or 6 residual blocks,i.e.model 1,2,3 and 4,respectively,and the computational cost and the amount of parameters of each model were recorded.The efficacy of the models were verified using test set,and the best DL model was screened.Results The computational cost of model 1,2,3 and 4 was 7 558.27,2 963.73,4 012.33 and 6 093.39 M,respectively,while the amount of parameters was 4.61,4.91,4.91 and 5.28 M,respectively.Model 2 had the least computational cost with parameters only slightly more than model 1.In test set,no significant difference of accuracy nor the area under the curve was found among 4 models(all P＞0.05).The sensitivity of model 2 was higher than that of model 3,while its specificity was lower only than that of model 3(both P＜0.05),hence model 2 was the best DL model.Conclusion Improved ResNet18 lightweight DL models could be used for automatically detecting GA based on ultrasonogram of MTP1,among which model 2 was the best one.

Objective Focusing on the problems of unbalanced nursing education resources and insufficient training,it discusses the path of nursing echelon training and nursing education resources optimization in secondary hospitals.Methods Based on nursing questionnaire,interview data and platform logs from 4 secondary hospitals in Hainan Province,Latent Dirichlet Allocation topic modeling,sentiment analysis and Social Network Analysis were used.Identify the core problems of educational resources and training system and propose optimization schemes.Results Topic modeling revealed that insufficient educational resources and information silos were major obstacles to shared platform development.Similarity analysis result showed high consistency in feedback regarding disconnected training content and single training models.Sentiment analysis result indicated that 50％of negative feedback focused on inadequate evaluation and incentive mechanisms.Social Network Analysis validated the significant role of multi-node collaboration in optimizing resource flows.Conclusion Building shared platforms,implementing differentiated training systems,and developing localized nursing faculty are critical pathways to achieving homogenized nursing education,providing valuable references for optimizing regional nursing education.

Objective To develop a toolkit suitable for assisting community health institutions in the early identification and inter-vention of mild cognitive impairment(MCI)among older adults.Methods A literature review was conducted to construct a draft of the identification and intervention toolkit.Tools with an expert approval rate above 70%were included after expert consultation.The final version of the toolkit was developed by integrating these tools with officially recommended tools in China.Results The expert consultation yielded an authority coefficient of 0.84.The finalized toolkit included the assessment tools of Mini-Mental State Examination,Montreal Cognitive Assessment,General Practitioner Assessment of Cognition,Cognitive Abilities Screening Instrument and Clock Drawing Test,and 18 intervention measures in-cluding pharmacological treatment,cognitive training and psychological interventions,etc.Conclusion The MCI Identification-Intervention Toolkit may serve as a reference for guiding the identification and inter-vention of MCI among older adults for community health institutions.

Objective To develop a value evaluation index system for health management and logistics positions in pub-lic hospitals,providing a research tool for future performance evaluations and salary reform.Methods Literature review,focus group interviews,and the point-factor method were employed to establish an initial index pool.A public hospital in Guangzhou was selected as a case study.Delphi expert consultation was then utilized to refine and finalize the indicators for the evaluation system.Results For the two rounds of Delphi expert consultation,the response rates were 87.5％and 100.0％,with the expert authority coefficient of 0.812.The Kendall's W coordination coefficients were 0.796 and 0.624 for the first round,and 0.747 and 0.918 for the second,(all＜0.001).The coefficients of variation ranged from 0.157 to 0.265 for the first round and from 0.108 to 0.230 for the second round.Experts provided four suggestions,leading to the removal of one secondary indicator and modifications to the definitions of three others.Finally,the evaluation system consists of 4 primary indicators and 12 secondary indicators.Conclusion This system can provide a research tool and reference for the follow-up performance evaluation,improve the salary reform of public hospitals,and promote the high-quality development of public hospitals.

Objective:To investigate clinical characteristics, treatment approaches, and prognosis of drug-induced hypersensitivity syndrome (DIHS) in children.Methods:A retrospective analysis was performed on clinical data from pediatric inpatients with DIHS in Department of Dermatology, Beijing Children's Hospital from 2009 to 2023. The clinical data included demographic characteristics, clinical manifestations, laboratory findings, treatment regimens, and outcomes.Results:A total of 103 children with DIHS were included, comprising 54 males (52.4%) and 49 females (47.6%), with ages ( M [ Q1, Q3]) of 2.3 (1.2, 4.5) years. Primary causative drugs were antibiotics (52 cases, 45.2%), antiepileptic drugs (41 cases, 35.7%), and nonsteroidal anti-inflammatory drugs (19 cases, 16.5%), with a median latency period of 12 days. All patients presented with rashes, including 72 (69.9%) with maculopapular rashes, 69 (67.0%) with edema (including 46 with facial edema). Lip involvement occurred in 25 cases (24.3%), and mucosal involvement was noted in 11 cases (10.7%). Additionally, 102 (99.0%) patients had fever, and 79 (76.7%) presented with lymphadenectasis. Eosinophilia was present in 64 cases (62.1%). Among 84 patients tested for atypical lymphocytes, 51 (60.7%) showed elevated percentages of atypical lymphocytes. Liver involvement was noted in 94 cases (91.3%), followed by pulmonary involvement in 31 (30.1%), gastrointestinal symptoms in 25 (24.3%), cardiac involvement in 14 (13.6%), renal involvement in 10 (9.7%), and pancreatic involvement in 7 cases (6.8%). Among 82 patients tested for blood immunocytes, 49 (59.8%) showed decreased percentages of B lymphocytes, and 69 (84.1%) showed decreased percentages of natural killer cells. Of 88 patients tested for serum immunoglobulins, 40 (45.5%) showed decreased IgA levels. Among 20 patients tested for serum cytokines, 15 (75.0%), 15 (75.0%), 13 (65.0%), and 12 (60.0%) showed elevated levels of interleukin (IL) -5, IL-6, IL-10, and interferon-γ, respectively. All patients received systemic glucocorticoid therapy, among whom 86 additionally received intravenous immunoglobulin therapy, 4 received Janus kinase inhibitors, and 3 received dupilumab. Five patients died, 9 developed hemophagocytic lymphohistiocytosis, 6 developed bronchiolitis obliterans, and 5 experienced long-term immune-related sequelae. Conclusions:Among these children with DIHS, antibiotics were the most common causative drugs, and the latency period could be shorter than 2 weeks. In addition to the common involvement of the liver and lungs, gastrointestinal and cardiac impairments were relatively frequent, while renal involvement was rare. Immunological features included decreased percentages of B lymphocytes and natural killer cells, reduced IgA levels, and elevated levels of cytokines such as IL-5, IL-6, IL-10, and interferon-γ.