Objective To explore the feasibility of a deep learning model based on early fundus lesions without manual segmentation in pan-retinal images for predicting diabetic kidney disease(DKD)and evaluating the enhancing effects of different binocular fusion strategies.Methods A retrospective cohort of 353 patients with type 2 diabetes mellitus(T2DM)admitted to the Endocrinology Department of Jiangdu People's Hospital Affiliated to Yangzhou University between December 2022 and March 2024 was analyzed.Patients were divided into DKD(n=114)and non-diabetic kidney disease(NDKD)(n=239)group based on the presence of DKD.First,a U-Net-based pre-trained automatic segmentation model was developed to process panoramic fundus retinal images.Subsequently,left and right eye deep learning models were constructed using ResNet152 under a five-fold cross-validation framework(70%training,30%validation).Three binocular fusion strategies were implemented:result fusion,feature fusion,and image fusion models.Model performance was evaluated using the area under the receiver operating characteristic(ROC)curve(AUC).DeLong test was used to compare AUC differences among models,while net reclassification index(NRI)and decision curve analysis(DCA)were used to assess clinical utility.Results Six prediction models were developed:clinical parameter model,left fundus model,right fundus model,binocular image fusion model,binocular result fusion model,and binocular feature fusion model.The Transformer-based binocular feature fusion model achieved the highest AUC in both training and validation sets(0.864 and 0.658,respectively).DeLong tests revealed significant AUC superiority of the Transformer model over the other five models in the training set(all P<0.001),though no significant differences were observed in the validation set(all P>0.05).NRI analysis showed negative values for all comparisons with the Transformer model(training set:-0.255,-0.244,-0.289,-0.426,-0.163;validation set:-0.060,-0.016,-0.028,-0.105,-0.033,respectively),indicating its optimal predictive performance.DCA further demonstrated greater net benefit for the Transformer-based fusion model.Conclusions The deep learning model constructed using early fundus lesions without manual segmentation in pan-retinal images can predict DKD.The Transformer-based fusion strategy present the best performance,providing a novel approach for further optimization and development of tools to predict DKD in the future.

OBJECTIVE To design and synthesize mono-carbonyl curcumin analogues(MCACs)and investigate the activities of them against breast cancer.METHODS The analogues F1,F2,and F3 were obtained by aldol condensation reaction,and their antitumor activities(including the activities of human breast cancer cell MCF-7 and human lung cancer cell A549)were detected by MTT assay[evaluated with half inhibitory concentration(IC50)].The results of MTT assay were compared with those of curcumin.Bioinformatics methods were used to collect the core targets of analogues F1,F2 and F3 acting on breast cancer,and then molecular docking verification was carried out.The cell experiments were conducted to investigate the effects of high,medium and low concentrations(16,8,4 μmol/L)of F1,F2 and F3 on the expression of the first core target protein as well as the effects of medium concentration of F1,F2 and F3 on the expression of cleaved-caspase-3.RESULTS Compared with curcumin,IC50 of analogues F1,F2 and F3 to A549 and MCF-7 cells(except for IC50 of analogue F2 to A549 cells)were decreased significantly(P＜0.05 or P＜0.01);among them,IC50of analogue F2 to MCF-7 cell was the lowest,being(9.67±1.27)μmol/L.Bioinformatics analysis showed that index of affinity of analogues F1,F2 and F3 with the first core target epidermal growth factor receptor(EGFR),protein kinase B(AKT)and AKT were 5.909 2,8.402 5 and 6.486 6,respectively;high concentration of F1 could significantly reduce the phosphorylation level of EGFR protein in MCF-7 cells(P＜0.01),while low,medium,and high concentrations of F2 and high concentration of F3 could significantly reduce the phosphorylation level of AKT protein in MCF-7 cells(P＜0.05 or P＜0.01).Medium concentration of F1,F2,and F3 could significantly increase the expression level of cleaved-caspase-3 protein in MCF-7 cells(P＜0.01).CONCLUSIONS Designed and synthesized MCACs F1,F2 and F3 all have good anti-breast cancer activity,and F2 has better anti-breast cancer activity.

Objective To explore the feasibility of a deep learning model based on early fundus lesions without manual segmentation in pan-retinal images for predicting diabetic kidney disease(DKD)and evaluating the enhancing effects of different binocular fusion strategies.Methods A retrospective cohort of 353 patients with type 2 diabetes mellitus(T2DM)admitted to the Endocrinology Department of Jiangdu People's Hospital Affiliated to Yangzhou University between December 2022 and March 2024 was analyzed.Patients were divided into DKD(n=114)and non-diabetic kidney disease(NDKD)(n=239)group based on the presence of DKD.First,a U-Net-based pre-trained automatic segmentation model was developed to process panoramic fundus retinal images.Subsequently,left and right eye deep learning models were constructed using ResNet152 under a five-fold cross-validation framework(70%training,30%validation).Three binocular fusion strategies were implemented:result fusion,feature fusion,and image fusion models.Model performance was evaluated using the area under the receiver operating characteristic(ROC)curve(AUC).DeLong test was used to compare AUC differences among models,while net reclassification index(NRI)and decision curve analysis(DCA)were used to assess clinical utility.Results Six prediction models were developed:clinical parameter model,left fundus model,right fundus model,binocular image fusion model,binocular result fusion model,and binocular feature fusion model.The Transformer-based binocular feature fusion model achieved the highest AUC in both training and validation sets(0.864 and 0.658,respectively).DeLong tests revealed significant AUC superiority of the Transformer model over the other five models in the training set(all P<0.001),though no significant differences were observed in the validation set(all P>0.05).NRI analysis showed negative values for all comparisons with the Transformer model(training set:-0.255,-0.244,-0.289,-0.426,-0.163;validation set:-0.060,-0.016,-0.028,-0.105,-0.033,respectively),indicating its optimal predictive performance.DCA further demonstrated greater net benefit for the Transformer-based fusion model.Conclusions The deep learning model constructed using early fundus lesions without manual segmentation in pan-retinal images can predict DKD.The Transformer-based fusion strategy present the best performance,providing a novel approach for further optimization and development of tools to predict DKD in the future.

Objective:To investigate the influence of maternal autoimmune diseases and anticoagulants, including low-molecular-weight heparin (LMWH) and aspirin, on the fetal fraction of maternal plasma cell-free DNA of non-invasive prenatal testing (NIPT).Methods:A prospective cohort study was conducted on women with singleton pregnancies receiving NIPT in the Nanjing Drum Tower Hospital from March 2021 to July 2022. NIPT was carried out using a polymerase chain reaction (PCR)-free amplification platform. In this study, four types of maternal autoimmune diseases, which were antiphospholipid syndrome, undifferentiated connective tissue disease, Sj?gren's syndrome, and systemic lupus erythematosus (SLE), and two anticoagulants, LMWH and aspirin, were studied. Univariate and multivariate linear regression models were used to analyze the factors influencing fetal fraction of maternal plasma cell-free DNA.Results:A total of 4 102 singleton pregnant women were enrolled in the prospective cohort, and 3 948 were finally included after excluding the cases with unclear dosing time of LMWH or aspirin, other autoimmune diseases, conceiving through ovulation induction alone, and having true positive or failed NIPT result. There were 96 cases with antiphospholipid syndrome, 35 with undifferentiated connective tissue disease, 34 with Sj?gren's syndrome, and 18 with SLE. A total of 108 patients only received LMWH treatment, 121 only received aspirin treatment, and 113 received both LMWH and aspirin treatment. Univariate linear regression analysis showed that maternal body mass index at blood collection ( B=－0.423), conceived by assisted reproductive technology ( B=－0.803), male fetus ( B=－0.458), undifferentiated connective tissue disease ( B=1.774), and SLE ( B=3.467) had influence on the fetal fraction (all P<0.05). Multivariate linear regression analysis showed that maternal body mass index at blood collection ( B=－0.415), conceived by assisted reproductive technology ( B=－0.585), male fetus ( B=－0.322), SLE ( B=3.347) and undifferentiated connective tissue disease ( B=1.336) were factors influencing fetal fraction (all P<0.05). Conclusions:Maternal use of LMWH or aspirin does not affect fetal fraction when performing NIPT on a PCR-free amplification platform, but undifferentiated connective tissue disease and SLE are the influencing factors. Therefore, pregnant women should be informed before the NIPT that the fetal fraction of maternal plasma cell-free DNA may be affected by maternal autoimmune diseases.

Background::Cerebrovascular disease (CVD) ranks among the foremost factors responsible for mortality on a global scale. The mortality patterns of CVDs and temporal trends in China need to be well-illustrated and updated.Methods::We collected mortality data on patients with CVD from Chinese Center for Disease Control and Prevention’s Disease Surveillance Points (CDC-DSP) system. The mortality of CVD in 2020 was described by age, sex, residence, and region. The temporal trend from 2013 to 2019 was evaluated using joinpoint regression, and estimated rates of decline were extrapolated until 2030 using time series models.Results::In 2019, the age-standardized mortality in China (ASMRC) per 100,000 individuals was 113.2. The ASMRC for males (137.7/10 5) and rural areas (123.0/10 5) were both higher when stratified by gender and urban/rural residence. The central region had the highest mortality (126.5/10 5), the western region had a slightly lower mortality (123.5/10 5), and the eastern region had the lowest mortality (97.3/10 5). The age-specific mortality showed an accelerated upward trend from aged 55-59 years, with maximum mortality observed in individuals over 85 years of age. The age-standardized mortality of CVD decreased by 2.43% (95% confidence interval, 1.02-3.81%) annually from 2013 to 2019. Notably, the age-specific mortality of CVD increased from 2013 to 2019 for the age group of over 85 years. In 2020, both the absolute number of CVD cases and the crude mortality of CVD have increased compared to their values in 2019. The estimated total deaths due to CVD were estimated to reach 2.3 million in 2025 and 2.4 million in 2030. Conclusion::The heightened focus on the burden of CVD among males, rural areas, the central and western of China, and individuals aged 75 years and above has emerged as a pivotal determinant in further decreasing mortalities, consequently presenting novel challenges to strategies for disease prevention and control.

Objective To investigate the effect of amlodipine and levamlodipine on the pharmacokinetics of lenvatinib and related mechanisms.Methods A total of 18 male Sprague-Dawley rats were randomly divided into lenvatinib(1.2 mg/kg)group,amlodipine(1.0 mg/kg)+lenvatinib group,and levamlodipine(0.5 mg/kg)+lenvatinib group,with 6 rats in each group.The rats were pretreated with 0.5%sodium carboxymethyl cellulose,amlodipine or levamlodipine by gavage for 8 days,and lenvatinib was given after the last intragastric administration.Blood samples were collected from the intraocular canthus venous plexus at the specified time points.Ultra-performance liquid chromatography-tandem mass spectrometry was used to measure the plasma concentration of lenvatinib in rats,and a non-compartment model was used to calculate pharmacokinetic parameters.RT-qPCR was used to measure the mRNA expression levels of cytochrome P450 3A1(CYP3A1),P-glycoprotein(P-gp),and breast cancer resistance protein(BCRP)in rat liver tissue.A one-way analysis of variance was used for comparison of normally distributed continuous data between multiple groups,and the Dunnett-t test was used for further comparison between two groups;the Kruskal-Wallis H test was used for comparison of non-normally distributed continuous data between groups.Results There were significant differences between the three groups in the area under the concentration-time curve AUC0-∞(F=4.567,P<0.05),clearance rate CLz/F(F=5.038,P<0.05),and peak concentration Cmax(F=11.667,P<0.01).Compared with the lenvatinib group,the amlodipine+lenvatinib group had an increase in AUC0-∞ by 36.1%(P<0.05),a reduction in CLz/F by 26.1%(P<0.05),and an increase in Cmax by 56.7%(P<0.01),and the levamlodipine+lenvatinib group had an increase in Cmax by 37.7%(P<0.05).RT-qPCR showed that there were significant differences in the mRNA expression levels of CYP3A1,P-gp,and BCRP between the three groups(F=10.160,5.350,and 5.237,all P<0.05),and compared with the lenvatinib group,the amlodipine+lenvatinib group had significant reductions in the mRNA expression levels of CYP3A1,P-gp,and BCRP in rat liver tissue(all P<0.05),while the levamlodipine+lenvatinib group had a significant reduction in the mRNA expression level of CYP3A1 in rat liver tissue(P<0.05).Conclusion Amlodipine can increase the in vivo exposure of lenvatinib possibly by inhibiting the mRNA expression of CYP3A1,P-gp,and BCRP in the liver,while levamlodipine only increases the peak concentration of lenvatinib.

Objective:To summarize the etiology, laboratory examinations and clinical features of rhabdomyolysis(RM)in children.Methods:The clinical data of children with RM admitted to the pediatric general ward at Third Hospital of Peking University from January 1st, 2010 to March 31st, 2021 were collected.The clinical characteristics, etiology distribution, laboratory examinations, treatments and prognosis of the children were analyzed.Results:A total of 24 children were included with 16 males and 8 females.The age ranged from 4 to 15 years old, with median age was 13years old.The etiology was exertional diseases in 14 cases(58.3%), non-exertional diseases in ten cases (41.7%, 7 cases of infection and 3 cases of other causes). The average age of exertional RM was(13.50±1.83)years, and that of non-exertional RM was(8.60±3.72)years.There was significant difference( t=3.848, P=0.002). The main clinical symptoms were muscle soreness, abnormal urine color and muscle weakness.Serum creatine kinase(CK)and serum myoglobin were significantly increased.The proportion of CK value moderate to severe increased of exercise RM children was significantly higher than that of non-exertional RM children( P=0.009). All children were treated with hydration and alkalization.Except for one case who died of critical primary disease, the other 23 children had good treatment response, and no rhabdomyolysis was found during the follow-up. Conclusion:The main causes of rhabdomyolysis in children are exercise and infection.Exertional RM is common in elder children.The increasing of CK level caused by exertional RM is more obvious.Active hydration and alkalization measures could avoid the occurrence of acute kidney injury.Most children with RM have good prognosis.

Objective: To explore the cross classification of tibial plateau fractures combined with avulsion fracture of medial femoral condyle. Methods: The 16 patients were retrospectively analysed as an observation group who had been treated at Department of Lower Limbs, Sichuan Orthopaedic Hospital for tibial plateau fracture and avulsion fracture of the medial condyle of the same femur from January 2015 to December 2018. They were 9 males and 7 females, aged from 27 to 78 years (average, 51.5 years). By the Schatzker classification, there were 2 cases of type Ⅰ, 8 cases of type Ⅱ, 3 cases of type Ⅲ and one case of type Ⅳ; by the cross classification, there were 10 cases of type Ⅰe, one case of type Ⅱe and 2 cases of type Ⅲe and 3 cases of type Ⅰv. The 332 patients were included as a control group who had been treated at the same department for simple tibial plateau fracture from January 2010 to December 2015. The imaging data of the 2 groups were compared to find out the charateristics of cross type distribution; the widened distance of the lateral tibial plateau, collapse depth of the tibial plateau, separation of avulsion fracture of medial femoral condyle, maximum clearance of the medial knee and eversion angle of the tibial plateau were investigated in the patients with cross types e and Ⅰv in the observation group. Results: In the observation group, the widened distance of the lateral tibial plateau averaged 7.18 mm, the collapse depth of the tibial plateau 8.74 mm, the separation of avulsion fracture of medial femoral condyle 3.44 mm, the maximum clearance of the medial knee 7.77 mm, and the eversion angle of the tibial plateau 87.03°. There was a significant difference in the cross type distribution of tibial plateau fractures between the 2 groups (P=0.002). In the observation group, the average widened distance of the lateral tibial platea for the 13 type e patients (8.49±4.26 mm)was significantly larger than that for the 3 type Ⅰv patients (1.51±2.11 mm) (t=2.706, P=0.017). Conclusion The cross classification of tibial plateau fractures with avulsion fracture of mediall condyle of the same femur may result mainly in types e and type Ⅰv. The widened distance of lateral tibial plateau helps distinction between types e and type Ⅰv.

Physical restraint is a complex and dynamic process in which nurses are the main decision makers and implementer. This article is to review the factors influencing the implementation of physical restraint in ICU nurses, and summarize from 4 aspects: nurses' own factors, environmental factors, promotion measures of rational use of physical restraint and ethical factors, so as to provide a basis for the effective clinical intervention measures in the future.

Objective To investigate the application value of limited liquid resuscitation combined with continuous renal replacement therapy (CRRT) in treatment of severe acute pancreatitis (SAP) companied with abdominal compartment syndrome (ACS).Methods The retrospective cohort study was adopted.The clinical data of 67 patients with SAP companied with ACS who were admitted to the Fuling Center Hospital of Chongqing from January 2005 to December 2014 were collected.Among 67 patients,33 receiving conventional liquid resuscitation between January 2005 and December 2010 were allocated into the control group and 34 receiving limited liquid resuscitation combined with CRRT between January 2011 and December 2014 were allocated into the observation group.Observation indicators included:(1) required fluid volume and time of negative fluid balance in the 2 groups;(2) changes of pathological and physiological indicators after treatment in the 2 groups;(3)outcomes and prognosis of patients in the 2 groups.Measurement data with normal distribution were represented as (x)±s and comparison between groups was analyzed using t test.Count data were analyzed using the chi-square test.Repeated measurement data were analyzed by repeated measures ANOVA.Results (1) Required fluid volume and time of negative fluid balance in the 2 groups:required fluid volumes at 6,24,48 and 72 hours after resuscitation were (2 449±339) mL,(4 820±757) mL,(9 428± 1 272) mL,(13 127± 1 565) mL in the control group and (2 360± 314) mL,(4 582±530) mL,(8 564± 970) mL,(11 470± 1 253) mL in the observation group,respectively,with a statistically significant difference in required fluid volume between the 2 groups (F=13.035,P＜0.05) and in required fluid volume at 48 and 72 hours between the 2 groups (t=3.132,4.794,P＜0.05).Time of negative fluid balance in the observation group and control group was (4.3± 1.7)days and (6.4 ±1.8)days,respectively,showing a statistically significant difference between the 2 groups (t =4.913,P＜0.05).(2) Changes of pathological and physiological indicators in the 2 groups after treatment:time factors:from 0 h to 168 h postoperatively,APACHE Ⅱ score,C-reaction protein (CRP),D-dimer,IAP,Bla and oxygenation index were changed from 20.9±4.1 to 13.9±2.6,from (167±39)mg/L to (55±17) mg/L,from (1 652±1 544) μg/L to (993±500)μg/L,from (23.4±3.4)cmH2O (1 cmH2O=0.098 kPa) to (21.4±2.9)cmH2O,from (4.6±1.6) mmol/L to (1.4±0.5)mmol/L,from (189±27) mmHg (1 mmHg =0.133 kPa) to (152±23)mmHg in the control group,and chaged from 21.5±5.1 to 11.0±2.8,from (168±36)mg/L to (44±19)mg/L,from (1 634±1 525) μg/L to (578±350) μg/L,from (23.2±2.5)cmH2O to (17.4±2.6)cmH2O,from (4.5±1.6)mmol/L to (0.8±0.3)mmol/L,from (188±26)mmHg to (196±25)mmHg in the observation group,respectively,showing gradual decreasing with time and statistically significant differences between the 2 groups (F =186.415,581.118,34.618,212.416,262.272,207.645,P＜0.05).Treatment factors:there were no significant differences in changing trends of APACHE Ⅱ score,D-dimer and Bla between the 2 groups (F=3.499,2.350,3.516,P＞0.05),and there were significant differences in changing trends of CRP,IAP and oxygenation index between the 2 groups (F=4.009,15.276,14.959,P＜0.05).Interaction effect between time factors and treatment factors:there were obviously interaction effects between time factors and treatment factors in APACHE Ⅱ score,CRP,IAP and oxygenation index (F=4.890,4.912,23.874,28.615,P＜0.05) and no interaction effects between time factors and treatment factors in D-dimer and Bla (F=2.803,1.920,P＞0.05).(3) Outcomes and prognosis of patients in the 2 groups:numbers of patients with surgery,local complications and infection and duration of hospital stay were 11,16,14,(46±17)days in the control group and 4,6,6,(36±14) days in the observation group,respectively,with statistically significant differences between the 2 groups (x2=4.484,7.221,4.910,t =2.433,P＜0.05).Mortality and hospital expenses were 24.2％ (8/33),(33± 18) x 104 yuan in the control group and 8.8％ (3/34),(27± 14)× 104 yuan in the observation group,respectively,with no statistically significant difference between the 2 groups (x2 =2.901,t =1.283,P＞ 0.05).Conclusion Limited liquid resuscitation combined with CRRT can effectively control IAP of patients with SAP companied with ACS and improve oxygenation index,meanwhile,it can also reduce number of patients with surgery,infection and local complications and duration of hospital stay.