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 detect the expression levels of long non-coding RNAs(lncRNA)in elderly patients with pulmonary tuberculosis(PTB)and those with non-tuberculous lung diseases(non-TB), and to assess the performance of these lncRNA in the differential diagnosis of PTB.Methods:A total of 300 elderly patients with suspected PTB were recruited from Beijing Chest Hospital between January 2024 and September 2024, and were further divided into the PTB group and the non-TB lung disease group based on the results of mycobacterium tuberculosis(MTB)pathogenicity testing.Peripheral blood mononuclear cells were isolated using a lymphocyte separation solution, and RNA was extracted using the TRIzol method.Nine lncRNAs, previously identified as differentially expressed in PTB through our group's microarray analysis, were selected and detected by real-time fluorescence quantitative polymerase chain reaction to evaluate the expression levels of these lncRNAs between the PTB and non-TB lung disease groups.The overall patients were randomly divided into training and validation sets in a 7∶3 ratio.Lasso regression was employed to select the characteristic variables, and a random forest algorithm was then used to construct the lncRNA diagnostic portfolio.Receiver operating characteristic(ROC)curves were generated to evaluate the diagnostic performance of individual lncRNAs and the combined panel in differentiating elderly patients with PTB from those with other non-TB lung diseases.Results:A total of 201 cases were included, with 105 confirmed elderly patients diagnosed with PTB(52.2%)and 96 elderly patients suffering from non-TB lung disease(47.8%).Compared to the elderly patients with non-TB lung disease, the expression levels of ENST00000417346.1, ENST00000620744.1, lncRNA PWP1, ENST00000583184.1, lncRNA ABHD17B, ENST00000607464.1, ENST00000516057.1, and NR_003000 were significantly downregulated in the PTB patients, whereas the expression level of lncRNA BCL2L10 was significantly upregulated in the PTB patients.ROC analysis revealed that the area under the curve(AUC)for each lncRNA ranged from 0.659 to 0.848.The diagnostic panel, which included NR_003000, ENST00000607464.1, ENST00000583184.1, and ENST00000620744.1 as determined by Lasso analysis, exhibited AUC values of 0.917 and 0.906 in the training and validation sets, respectively.The performance of this panel was superior to that of each individual lncRNA.Conclusions:The random forest model, which incorporates NR_003000, ENST00000607464.1, ENST00000583184.1, and ENST00000620744.1, demonstrates potential in differentiating between PTB and non-TB lung diseases.

Objective To explore the roles of the mechanoreceptor Yes-associated protein 1(YAP1)and piezo type mechanosensitive ion channel component 2(PIEZO2)in mechanotransduction in mouse bladder urothelial cells.Methods Mouse bladder urothelial cells were subjected to mechanical stretching using the FX-6000T cell stretching system and treated with the YAP1-specific inhibitor verteporfin(VP).The expressions of PIEZO2,YAP1 and connective tissue growth factor(CTGF)at the mRNA and protein levels,as well as changes in cellular adenosine triphosphatase(ATP)concentration,were detected using reverse transcription quantitative PCR(RT-qPCR)and Western blotting(WB).Results After stretching stimulation,under the fluorescence microscope,it was observed that the diameter length of the stretched cells were longer than that before stretching,and the difference was statistically significant(P＜0.05).The expressions of YAP1,PIEZO2 and CTGF at the mRNA and protein levels were increased in the stretched group compared to those of the non-stretched group(P＜0.05).VP effectively reduced the expressions of YAP1,PIEZO2 and CTGF at the mRNA and protein levels after stretching stimulation(P＜0.05).Stretching stimulation significantly increased the intracellular ATP concentration,while VP was able to inhibit the increase in ATP concentration,with a statistically significant difference(P＜0.000 1).Conclusion Stretching stimulation increased the expressions of YAP1 and PIEZO2 in bladder urothelial cells and promoted the release of ATP;verteporfin inhibited the increase in YAP1 activity and the overexpression of PIEZO2 caused by stretching,thereby reducing the release of ATP.It is suggested that mouse bladder urothelial cells may primarily sense mechanical signals through the YAP1-PIEZO2-ATP pathway.

Objective: To analyze the association of participation in non-sports extracurricular tutoring classes with the prevalence of screening myopia, axial length (AL) and axial length to corneal radius ratio (AL/CR) among primary school students, so as to provide evidences for formulating myopia prevention and control policies. Methods: In December 2024, combination of convenience and cluster sampling method was used to select 2 273 students from two primary schools in Hefei City, Anhui Province. Ophthalmic examinations and questionnaire surveys were conducted to obtain information on myopia, AL, AL/CR and participation in various types of extracurricular tutoring. A binary Logistic regression model was used to analyze the association between non-sports tutoring and screening myopia, and multiple linear regression models were used to examine the associations between non-sports tutoring and AL and AL/CR. Results: Among the surveyed students, the participation rate in non-sports extracurricular tutoring classes was 64.9% , and the overall prevalence of screening myopia was 39.1%. The average AL and AL/CR were (23.60± 1.01 ) mm and (3.00±0.12), respectively. Univariate analysis showed that students who attended non-sports, music, or academic tutoring classes for ≥2 h per week had higher risks of screening myopia and greater AL/CR values than non-participants (screening myopia: OR =1.38, 1.82, 1.55; AL/CR: β =0.01, 0.03, 0.03; all P <0.05). After adjusting for sex, grade, and participation in sports tutoring, multivariate analysis indicated that participation in non-sports and musical instrument tutoring classes for ≥2 h per week remained significantly associated with higher risks of screening myopia ( OR =1.26, 1.49, both P <0.05). Multiple linear regression showed that participation in musical instrument tutoring for ≥2 h per week was positively correlated with AL ( β=0.14, P < 0.05). Conclusions Participation in non-sports extracurricular tutoring is common among primary school students. Attending non-sports tutoring classes for ≥2 h per week increases the risk of screening myopia.

Objective To develop and validate risk prediction models utilizing five machine learning algorithms for assessing postoperative pulmonary infection(PPI)risk in lung cancer patients undergoing grade Ⅳ thoracoscopic surgery.Methods A retrospective cohort study included 2,380 lung cancer patients who underwent grade Ⅳ thoracoscopic surgery at a tertiary hospital in Shanghai(January 2022 to June 2024).Patients were stratified into training(n=1,665)and validation(n=715)cohorts.Five machine learning algorithms—Logistic regression(LR),artificial neural network(ANN),support vector machine(S VM),random forest(RF),and extreme gradient boosting(XGB)—were employed to construct predictive models.A nomogram was developed for clinical utility.Results Among 2,380 patients,226(9.5％)developed PPI.The Least Absolute Shrinkage and Se-lection Operator(LASSO)regression identified eight predictive variables:daily cigarette consumption,diabetes history,preoperative diffusing capacity,maximal tumor diameter,24-hour postoperative chest drainage volume,perioperative oral nutritional supplementation(ONS),postoperative urinary cathe-terization,and intraoperative pleural adhesion severity.All models demonstrated robust discrimina-tion,with area under the curve(AUC)values ranging from 0.862 to 0.947.The XGB model a-chieved superior performance(AUC=0.947,95％CI,0.937 to 0.962),followed closely by the LR model(AUC=0.926,95％CI,0.918 to 0.933).Conclusion Machine learning-based algo-rithms models effectively stratify PPI risk in lung cancer patients following grade Ⅳ thoracoscopic surgery.The derived nomogram provides a practical tool for perioperative risk management by healthcare providers.

Objective To analyze the clinical manifestations and diagnostic characteristics of maternal cardiac arrest(MCA)patients,summarize warning symptoms,and provide a basis for the prevention,early identification,and intervention of this event.Methods A retrospective study was conducted.Clinical data of 15 pregnant or postpartum women who experienced cardiac arrest or were transferred to Peking University Third Hospital due to cardiac arrest from January 2012 to December 2023 were collected,including age,comorbidities,causes of MCA,warning symptoms,parameters related to cardiopulmonary resuscitation(CPR,such as time and location of cardiac arrest,initial heart rhythm,electrical defibrillation,medication use,duration of CPR,restoration of spontaneous circulation,and etc.),parameters related to perimortem cesarean delivery(time interval between fetal delivery and cardiac arrest,delivery of infants),clinical outcomes(such as cerebral performance category,length of hospital stay),and multidisciplinary treatment.The main clinical endpoint was patient survival at discharge.The clinical and diagnostic characteristics of MCA were descriptively analyzed,and compare the differences in age,duration of CPR,shockable rhythm,proportion of patients with cardiac arrest occurring prenatally,and length of hospital stay among patients with different prognoses of MCA.Results The median age of 15 MCA patients was 31(30,33)years.A total of 10 cases(66.7%)of patients had concurrent obstetric diseases.Six cases(40.0%)of patients had MCA due to postpartum hemorrhage,which was the most common cause,followed by acute heart failure[2 cases(13.3%)]and pulmonary embolism[2 cases(13.3%)].A total of 12 cases(80.0%)of MCA patients showed obvious warning symptoms before cardiac arrest,with the most common warning symptoms including dyspnea(3 cases,20.0%),consciousness disorders(3 cases,20.0%),and vaginal bleeding(3 cases,20.0%).A total of 12 cases(80.0%)of MCA occurred in late pregnancy.All 15 MCA events occurred within the hospital,with an average of 5(4-6)departments involved in the CPR process.Four patients(26.7%)had an initial heart rhythm of ventricular fibrillation and all received defibrillation therapy,9 patients(75.0%)were treated with adrenaline.Two patients who were over 20 weeks pregnant underwent manual left uterine thrusting intervention during CPR,and both patients were successfully discharged.A total of 12 patients(80.0%)received CPR,of which 7 patients(58.3%)were discharged with good neurological function.A total of 11 patients underwent cesarean section surgery,and the survival rate of fetuses delivered by cesarean section was 83.3%.The duration of CPR in survivors was significantly shorter than that in deceased patients[minutes:7.0(2.0,23.3)vs.144.0(90.0,190.5),P<0.05].Conclusion Postpartum hemorrhage is the most common cause of MCA,and after active treatment,over 50%of patients can achieve good neurological function and be discharged.Comprehensive and effective CPR,strengthened coordination among multidisciplinary teams,and early identification and warning of symptoms are key strategies for improving the prognosis of MCA patients.

Objective:To evaluate the effectiveness and safety of allogeneic hematopoietic stem cell transplantation (allo-HSCT) in children with stage 4/M neuroblastoma (NB).Methods:This study was a prospective, single-arm, multicenter clinical trial conducted by Sun Yat-sen Memorial Hospital, Children′s Hospital of Nanjing Medical University, Children′s Hospital, Zhejiang University School of Medicine, the First Affiliated Hospital of Guangxi Medical University, Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine. From March, 2019 to August, 2023, 25 children with confirmed with stage 4/M NB and received allo-HSCT were enrolled. The patients received either unrelated cord blood transplantation (UCBT) or peripheral blood stem cell transplantation (PBSCT). Conditioning regimens for UCBT was fludarabine+busulfan+cyclophosphamide+topotecan, and for PBSCT was fludarabine+busulfan+melphalan+thiotepa+antithymocyte globulin, respectively. Until the last follow-up date of September, 2023, the overall survival (OS) rate and event free survival (EFS) rate were analyzed to evaluate efficacy. The engraftment rate and transplant-related complications were statistically assessed to evaluate safety. Survival analysis was performed using the Kaplan-Meier method.Results:Of the 25 patients, there were 15 males and 10 females. The age at transplantation was 5.7 (3.8, 7.3) years. The engraft rate was 100%, with recovery time of neutrophil as 15.7 (12.5, 17.0) d, and the recovery time of platelets as 33.5 (18.0, 48.0) d. Seventeen of the 25 children (68%) developed acute graft versus host disease (aGVHD), occurred at 18.0 (13.0, 22.5) d after transplantation, including 13 of grade Ⅲ-Ⅳ cases. The main sites of aGVHD were skin and intestinal tract. After treatment, 13 cases improved, 4 patients developed chronic graft-versus-host disease (cGVHD). After allo-HSCT, 14 children received maintenance therapy. Twenty of the 25 patients survived, the 2-year cumulative OS rate was (80±9)%, and 2-year EFS rate was (56±11)%. Nine cases (36%) relapsed, the time from allo-HSCT to disease relapse was 10.9 (5.5, 16.0) months. Five cases (20%) died. The hematopoietic stem cell transplantation associated mortality rate was 4% (1/25).The 2-year OS rate of patients who had partial remission prior to allo-HSCT was significant lower than those who had complete remission prior to allo-HSCT ((33±25)% vs. 100%, P=0.037). Conclusion:allo-HSCT is an effective treatment for patients with stage 4/M NB.

Objective:To investigate the role of mechanosensor Yes-associated protein 1 (YAP1) in urothelial cells in inducing bladder dysfunction in a partial bladder outlet obstruction (pBOO) model.Methods:Ten female C57BL/6 mice were included in this study and randomly divided into pBOO and sham groups based on body weight using a stratified pairing method, with 5 mice in each group. The pBOO group underwent proximal urethral ligation surgery, while the sham group underwent a sham operation. Two weeks after surgery, the urinary pattern was analyzed using the urine spot test. The significant increase in urine spot numbers indicated the successful establishment of the pBOO model. The mice were then sacrificed, and bladder tissues were weighed and stained with hematoxylin and eosin (HE) to observe morphological changes. The bladder urothelial layer was further isolated, and total cell proteins were extracted to detect the expression levels of YAP1 protein using Western blotting. Mouse immortalized bladder urothelial cells were divided into three experimental groups: the negative control (NC) group, which was treated with YAP1-NC lentivirus; the overexpression (OE) group, which was treated with YAP1-OE lentivirus to induce YAP1 protein overexpression; and the verteporfin treatment (VP) group, which was treated with verteporfin on the basis of the OE group. Real-time quantitative PCR and Western blotting were used to verify the transcription and expression levels of YAP1 protein, the co-transcriptional activator TEAD4 protein, and the phosphorylated protein DRP1-616 (at serine 616) of dynamin-related protein 1 (DRP1). An ATP detection kit was used to measure the ATP release concentration in the NC, OE, and VP groups. The interaction between YAP1 and TEAD4 was investigated using co-immunoprecipitation, and the expression of the mitochondrial marker translocase of the outer mitochondrial membrane 20 (Tom20) was observed using immunofluorescence staining.Results:The results of the urine spot test showed that the number of urine spots on the filter paper in the pBOO group was higher than that in the sham group within 6 hours [(283.0±9.1) spots vs. (3.7±0.3) spots, P<0.01], and the urine spots were scattered. The bladder wet weight in the pBOO group was significantly higher than that in the sham group [(105.70±6.84) mg vs. (22.33±1.20) mg, P<0.01]. Histological observations revealed reduced bladder mucosal folds and increased detrusor muscle thickness in the pBOO group. The expression of YAP1 protein in the bladder urothelial cells of the pBOO group was significantly upregulated compared to the sham group [(1.26±0.08) vs. (0.50±0.04), P<0.01]. In vitro experiments showed that compared to the NC group, the OE group had significantly increased expression of DRP1-616 [(0.94±0.05) vs. (0.33±0.01), P<0.01] and higher ATP release concentration [(24.45±0.16) μmol/mg vs. (19.67±0.42) μmol/mg, P<0.01]. In contrast, the VP group had significantly decreased expression of DRP1-616 [(0.29±0.04) vs. (0.94±0.05), P<0.01] and lower ATP release concentration [(10.55±0.01) μmol/mg vs. (24.45±0.16) μmol/mg, P<0.01] compared to the OE group. Co-immunoprecipitation experiments using YAP1 and TEAD4 antibodies showed that YAP1 and TEAD4 proteins could interact and form a transcriptional complex to regulate ATP release. Immunofluorescence staining revealed increased expression of Tom20 in the OE group compared to the NC group [(104.20±3.28) vs. (74.51±3.87), P<0.01]. Conclusions:In the pBOO-induced bladder dysfunction model, YAP1 is highly expressed in urothelial cells. YAP1 forms a transcriptional complex with TEAD4 to regulate ATP release by promoting mitochondrial fission via DRP1-616 expression, which is a key mechanism underlying pBOO-induced bladder dysfunction.

Objective:To analyze the epidemiological characteristics and infection sources of cholera in China from 2005 to 2024.Methods:A total of 2 066 cholera cases were included in the study, which were obtained from the China Disease Control and Prevention Information System (CDPCIS) of China CDC. The information on cholera clusters was downloaded from the National Public Health Emergency Event Surveillance System (PHEESS) of China CDC. A total of 128 cholera clusters were included and analyzed in this study. The epidemiological characteristics and infection sources of cholera were analyzed. The Jointpoint model was applied to analyze the incidence trend, and annual percentage change (APC) was also quantified.Results:From 2005 to 2024, a total of 2 066 cholera cases were reported, with an average of 103 cases reported annually. Specifically, the incidence showed a marked downward trend from 2004 to 2015 ( APC=-26.78%, P=0.006). During 2015-2024, the disease remained at low endemic levels, with an average of 18 reported cases annually ( APC=-2.68%, P=0.807). Cholera peak season was from May to October. A total of 24 provinces reported cholera cases, which were mainly distributed in Zhejiang, Fujian, Beijing, Jiangsu, Anhui, Guangdong, and Hainan provinces, accounting for 78.03% of the total cases. Pathogen surveillance indicated an alternating prevalence of Vibrio cholerae serogroups O1 and O139 among laboratory-confirmed cases between 2005 and 2024. There was a disparity in the dominant serogroup of Vibrio cholerae by region. The results from 128 cholera clusters indicated that cholera outbreaks frequently occurred in rural banquets (64.84%), followed by regular restaurants (13.28%). Among these, 63 clusters (49.22%) with identified infection sources indicated that foodborne transmission (95.24%) was the primary mode of cholera transmission, which mainly through seafood and aquatic products, such as soft-shelled turtles, shrimp and shellfish. The characteristics of cholera clusters caused by Vibrio cholerae serogroups O1 and O139 showed statistically significant differences in scale, attack rate, place of residence, setting, and infection source ( P<0.05). Conclusion:Cholera incidence has remained consistently low since 2015 in China, mainly in sporadic cases. Rural gatherings (e.g., wedding banquets) are the main settings for cholera clusters. The main infection sources are predominantly caused by cross-contamination due to improper processing practices of aquatic products, such as soft-shelled turtles.

Objective To explore the effectiveness and practical pathways of integrating narrative education into higher vocational nursing courses for cultivating students' humanistic care ability.Methods A total of 156 nursing students from a higher vocational college were randomly divided into an observation group(narrative education intervention)and a control group(traditional teaching).Quantitative evaluations were conducted using the Caring Ability Inventory(CAI)and a self-developed Humanistic Care Quality Scale,supplemented by semi-structured interviews for qualitative thematic analysis.Results The ob-servation group demonstrated significantly higher total caring ability scores(180.1±10.9 vs 145.2±14.3),humanistic care quality(97.5±3.8 vs 77.7±5.9),and course satisfaction(9.5±0.6 vs 7.0±1.3)compared to the control group(P＜0.01).Qualitative findings revealed that narrative education enhanced professional beliefs and practical skills through empathy training,reflective writing,and team collaboration.Conclusion Narrative education systematically improves the humanistic care ability and professional of higher vocational nursing students,providing empirical support for nursing education reform.