OBJECTIVE: To identify the optimal machine learning algorithm for predicting post-stroke epilepsy (PSE) within one year following acute stroke, establish a nomogram model based on this algorithm, and perform external validation to achieve accurate prediction of secondary epilepsy. METHODS: A total of 870 acute stroke patients admitted to the emergency department of Xiang'an Hospital of Xiamen University from June 2019 to June 2023 were enrolled for model development (model group). An external validation cohort of 435 acute stroke patients admitted to the Fifth Hospital of Xiamen during the same period was used to validate the machine learning algorithms and nomogram model. Patients were classified into control and epilepsy groups based on the development of PSE within one year. Clinical and laboratory data, including baseline characteristics, stroke location, vascular status, complications, hematologic parameters, and National Institutes of Health Stroke Scale (NIHSS) score, were collected for analysis. Nine machine learning algorithms such as logistic regression, CN2 rule induction, K-nearest neighbors, adaptive boosting, random forest, gradient boosting, support vector machine, naive Bayes, and neural network were applied to evaluate predictive performance. The area under the curve (AUC) of receiver operator characteristic curve (ROC curve) was used to identify the optimal algorithm. Logistic regression was used to screen risk factors for PSE, and the top 10 predictors were selected to construct the nomogram model. The predictive performance of the model was evaluated using the ROC curve in both the model and validation groups. RESULTS: Among the 870 patients in the model group, 29 developed PSE within one year. Among the nine algorithms tested, logistic regression demonstrated the best performance and generalizability, with an AUC of 0.923. Univariate logistic regression identified several risk factors for PSE, including platelet count, white blood cell count, red blood cell count, glycated hemoglobin (HbA1c), C-reactive protein (CRP), triglycerides, high-density lipoprotein (HDL), aspartate aminotransferase (AST), alanine aminotransferase (ALT), activated partial thromboplastin time (APTT), thrombin time, D-dimer, fibrinogen, creatine kinase (CK), creatine kinase-MB (CK-MB), lactate dehydrogenase (LDH), serum sodium, lactic acid, anion gap, NIHSS score, brain herniation, periventricular stroke, and carotid artery plaque. Further multivariate logistic regression analysis showed that white blood cell count, HDL, fibrinogen, lactic acid and brain herniation were independent risk factors [odds ratio (OR) were 1.837, 198.039, 47.025, 11.559, 70.722, respectively, all P < 0.05]. In the external validation group, univariate logistic regression analysis showed that platelet count, white blood cell count, CRP, triacylglycerol, APTT, D-dimer, fibrinogen, CK, CK-MB, LDH, NIHSS score, and cerebral herniation were risk factors for PSE one year after acute stroke. Further multiple logistic regression analysis showed that APTT and cerebral herniation were independent predictors (OR were 0.587 and 116.193, respectively, both P < 0.05). The nomogram model, constructed using 10 key variables-brain herniation, periventricular stroke, carotid artery plaque, white blood cell count, triglycerides, thrombin time, D-dimer, serum sodium, lactic acid, and NIHSS score-achieved an AUC of 0.908 in the model group and 0.864 in the external validation group. CONCLUSIONS The logistic regression-based prediction model for epilepsy one year after acute stroke, developed using machine learning algorithms, showed optimal predictive performance. The nomogram model based on the logistic regression-derived predictors showed strong discriminative power and was successfully validated externally, suggesting favorable clinical applicability and generalizability.
Humans ; Machine Learning ; Stroke/complications* ; Nomograms ; Epilepsy/etiology* ; Algorithms ; Male ; Female ; Logistic Models ; Middle Aged ; Aged ; Risk Factors ; Bayes Theorem

Objective:To construct univariate and multivariate relapse free survival (RFS) prediction models for breast cancer patients with diabetes mellitus type 2 (T2DM) and to compare and select the model with higher predictive performance.Methods:A total of 912 breast cancer patients treated at the First Affiliated Hospital of Dalian Medical University from January 2010 to December 2016 were included, of which 202 patients had T2DM and 710 patients did not. Kaplan-Meier survival curve was drawn based on whether patients had T2DM, and log-rank test was performed based on whether patients had T2DM. All patients were randomly divided into a training set ( n=640) and a validation set ( n=272) at a ratio of 7∶3. Univariate and multivariate Cox proportional risk regression models were used to analyze RFS in breast cancer patients with the survival package. The "rms" package was employed to construct univariate and multivariate RFS prediction models for breast cancer patients with T2DM. Clinical decision curves and calibration curves were used to validate the models. The receiver operator characteristic (ROC) curve was used to compare and analyze the prediction performance of the two models. Results:There were no statistically significant differences between the training set and the validation set patients in terms of age, T2DM, surgical approach, axillary management methods, T stage, N stage, molecular sub-type, estrogen receptor (ER) 1, ER2, progesterone receptor (PR) , ER and PR consistency, Ki67, human epidermal growth factor receptor 2 (HER2) (all P>0.05) . There was a statistically significant difference in histological grade ( χ2=7.59, P=0.022) . Survival analysis showed that the 5-year RFS rate was 83.7% in patients with T2DM and 92.3% in patients without T2DM ( χ2=16.61, P<0.001) . Univariate analysis revealed that age ( HR=1.04, 95% CI: 1.03-1.06, P<0.001) , T2DM ( HR=2.31, 95% CI: 1.49-3.55, P<0.001) , surgical approach ( HR=2.39, 95% CI: 1.20-4.77, P=0.013) , axillary management methods ( HR=2.62, 95% CI: 1.72-3.98, P<0.001) , T stage (T 2: HR=2.13, 95% CI: 1.36-3.31, P<0.001; T 3: HR=6.90, 95% CI: 3.35-14.22, P<0.001) , N stage (N 2: HR=3.87, 95% CI: 2.12-7.07, P<0.001; N 3: HR=8.61, 95% CI: 4.71-15.75, P<0.001) , molecular sub-type (Luminal B: HR=2.74, 95% CI: 1.17-6.36, P=0.019; HER2 +: HR=3.64, 95% CI: 1.38-9.58, P=0.009; TNBC: HR=4.40, 95% CI: 1.71-11.34, P=0.002) , ER1 (>10%: HR=0.57, 95% CI: 0.37-0.90, P=0.016) , ER2 ( HR=0.57, 95% CI: 0.37-0.89, P=0.015) , and PR ( HR=0.56, 95% CI: 0.37-0.86, P=0.008) were all factors influencing RFS in breast cancer patients. Multivariate analysis demonstrated that age ( HR=1.04, 95% CI: 1.02-1.06, P<0.001) , T2DM ( HR=1.82, 95% CI: 1.16-2.85, P=0.009) , T stage (T 2: HR=1.60, 95% CI: 1.01-2.54, P=0.046; T 3: HR=2.64, 95% CI: 1.22-5.72, P=0.014) , N stage (N 2: HR=3.72, 95% CI: 2.01-6.88, P<0.001; N 3: HR=5.34, 95% CI: 2.78-10.25, P<0.001) , and ER1 (>10%: HR=0.63, 95% CI: 0.39-0.99, P=0.046) were independent factors influencing RFS in breast cancer patients. Based on the 10 and 5 variables with P<0.05 in the univariate and multivariate analyses respectively, the nomograms of the univariate and multivariate prediction models were constructed to evaluate the influence of factors such as T2DM on the postoperative RFS of breast cancer patients. Clinical decision curves and calibration curves indicated that both models had high predictive value for RFS in breast cancer patients, and the predictive results were highly consistent with the actual observed results. ROC curve analysis showed that there was no statistically significant difference in the area under the curve (AUC) of the two models for predicting the RFS rates of breast cancer patients in the training set and validation set at 36, 60, and 84 months (all P>0.05) , indicating that the predictive efficacy of the two models was comparable. The multivariate model is more suitable for clinical application because it uses fewer variables. Conclusions:Breast cancer patients with T2DM have poorer prognosis. Age, T2DM, T stage, N stage, and ER1 are independent factors influencing postoperative RFS in breast cancer patients. The multi-factor prediction model of RFS in breast cancer patients based on T2DM is more suitable for clinical application due to its higher predictive efficacy and fewer variables.

Objective:To analyze mortality risk factors in patients with severe pneumonia and investigate their varying influences across different time periods.Methods:A total of 134 patients with severe pneumonia admitted to the Emergency Department of Xiang’an Hospital, Xiamen University, between June 2019 and February 2020 were enrolled. All patients were treated in the EICU and followed up for four years. Based on outcomes, they were categorized into a death group ( n=77) and a survival group ( n=57). COX regression analysis was employed to identify mortality risk factors at different time points, while logistic regression analysis was used to assess risk factors influencing mortality during hospitalization, ICU stay, 1-month, and 1-year follow-up periods. Results:Mortality rates were 11.9% ( n=16) during ICU admission, 20.8% ( n=28) during hospitalization, 16.4% ( n=22) within 1 month, and 31.3% ( n=42) within 1 year. By the end of the follow-up, 57.4% ( n=77) of patients had died. Ten mortality risk factors were identified, with the number increasing over time. During ICU admission and hospitalization, significant risk factors included total bilirubin levels, APACHE-II score, invasive ventilation, ARDS, and vasopressor use in the ICU. One-month mortality risk additionally involved bacterial infection. One-year mortality risk further incorporated advanced age and chronic heart failure. By the end of follow-up, acute kidney injury (AKI) during ICU admission also emerged as a contributing factor, while higher body weight was identified as a protective factor. Conclusions:The number of mortality risk factors in severe pneumonia patients increases progressively over time. Early-stage factors during hospitalization and ICU admission exert a stronger impact on short-term mortality, whereas bacterial infection, advanced age, and chronic heart failure become increasingly significant in later stages. These findings highlight the dynamic nature of risk factors and underscore the importance of tailored monitoring and intervention strategies at different disease phases.

Background Ozone (O₃) pollution has gradually become a primary problem of air pollution in recent years. Conducting epidemiological studies on the correlation between O₃ concentration variation and risk of cardiovascular and cerebrovascular diseases can provide reference data for O₃ risk assessment and related policy making. Objective To quantitatively evaluate the effects of O₃ exposure on mortalities of cardiovascular and cerebrovascular diseases among residents in Minhang District, Shanghai. Methods Data of mortalities of cardiovascular and cerebrovascular diseases, air pollutants, and meteorological factors in Minhang District of Shanghai from January 1, 2016 to December 31, 2021 were collected. Associations between O₃ concentration and the mortalities due to total cardiovascular and cerebrovascular diseases, coronary heart disease, and stroke were analyzed by generalized additive models with a quasi Poisson distribution with different lag patterns, such as current day effect (lag0), single-day lag effects (lag1-lag3), and cumulative lag effects (lag01-lag03). The subgroup analyses of age, sex, and season were conducted. Furthermore, temperature was divided into low, middle, and high levels based on the 25th percentile (P₂₅) and the 75th percentile (P₇₅) to perform hierarchical analyses. Increased excess risks (ER) of death from target diseases caused by a 10 µg·m⁻³ increase in daily maximum 8 h concentration of O₃ (O₃-8 h) and their 95% confidence intervals (CI) were used to indicate the effects of O₃. Results The associations between O₃ and the risks of death from cardiovascular and cerebrovascular diseases were statistically significant at lag2, lag3, lag02, and lag03 (P<0.05), with the greatest effect size observed at lag03. The ER values of death from cardiovascular and cerebrovascular diseases in general population, male residents, and people aged 65 years and older, from coronary heart disease in male residents, and from stroke in general population increased by 1.02% (95%CI: 0.36%, 1.69%), 1.40% (95%CI: 0.47%, 2.34%), 0.87% (95%CI: 0.19%, 1.55%), 1.96% (95%CI: 0.49%, 3.44%), and 1.02% (95%CI: 0.07%, 1.98%) for a 10 µg·m⁻³ increase in O₃-8 h concentration at lag03, respectively. During the warm season (from April 1 to September 30), the ER values of death from cardiovascular and cerebrovascular diseases and coronary heart disease per 10 µg·m⁻³ increase in O₃ were 1.18% (95%CI: 0.33%, 3.33%) and 2.69% (95%CI: 0.39%, 5.03%), while the O₃ effect was only statistically significant on cardiovascular and cerebrovascular diseases during the cold season (from October 1 to March 31 next year). At the middle and high temperature levels, the ER values of death from cardiovascular and cerebrovascular diseases increased by 1.63% (95%CI: 0.32%, 2.96%) and 1.14% (95%CI: 0.17%, 2.12%) respectively. The two-pollutant models showed similar results after including other pollutants (carbon monoxide, nitrogen dioxide, sulfur dioxide, fine particulate matter, or inhalable particulate matter). Conclusion Ambient O₃ pollution may increase the mortality risks of cardiovascular and cerebrovascular diseases, coronary heart disease, and stroke in Minhang District of Shanghai.

Objective To investigate the differences in blood pressure phenotypes,renal pathological changes,and related pathogenic pathways in genetically diverse hypertensive mice obtained from 13 strains.Methods The genotypes of Cckbr+/+,Cckbr+/-and Cckbr-/-were obtained by hybridization of 13 strains of genetically diverse mice with Cckbr-/-mice.Blood pressure was measured with a noninvasive blood pressure analysis system(BP-2000).The expression of CCKBR protein in mouse kidney tissue was detected by Western Blot,and the pathological changes in mouse kidney tissue were detected by hematoxylin-eosin(HE)staining and immunohistochemistry(IHC).The pathogenic pathways related to essential hypertension were screened by RNA sequencing.Results In three specific mouse strains(A/J,LOT,and FIM),the systolic blood pressure(SBP)was significantly different between the Cckbr-/-and Cckbr+/+groups.HE staining and IHC showed that hypertension caused a certain degree of renal injury in the mice.Gene Ontology(GO)and pathway enrichment analysis showed that differentially expressed genes were enriched in metabolic processes and circadian rhythm regulation.Conclusions Genetically diverse mice can effectively simulate the genetic background of the population and provide a new resource for studying the pathogenic genes related to essential hypertension.

Objective:To construct an evidence-based prediction model for early recurrence after surgery of hepatocellular carcinoma (HCC) based on Meta-analysis and to do external validation study.Methods:The literatures in Chinese National Knowledge Infrastructure, Wanfang, VIP, Chinese Science Citation Database (CSCD), Chinese Social Science Citation System (CCSCI), PubMed, Web of Science and IEEE databases between January 2019 and December 2023 were searched based on the subject words. According to the inclusion and exclusion criteria, 9 literatures were included to screen the risk factors affecting the early recurrence of HCC. When the same risk factor was found in ≥5 included literatures, Meta-analysis was performed by using Review Manager 5.4.1 software. External validation data were collected from 401 patients with primary HCC who underwent surgery in Liaoning Cancer Hospital between March 2014 and March 2017. The patients were divided into early recurrence group (176 cases) and early non-recurrence group (225 cases) according to whether they relapsed 2 years after surgery. The OR values of all risk factors obtained in the Meta-analysis were converted into modeling, and postoperative early recurrence rate of HCC in the Meta-analysis was used to calculate β 0, and finally the logistic model was obtained. The OR value was incorporated into the logit (P) model, and the morbidity (P) of the external validation data was calculated. Taking the recurrence 2 years after surgery or not as the dependent variable and P as the independent variable, the receiver operating characteristic (ROC) curve was drawn to calculate the area under the curve (AUC). Results:A total of 8 risk factors for early HCC recurrence were screened out from 9 literatures (x 1: alpha-fetoprotein ≥ 400 ng/ml; x 2: tumor number ≥ 2; x 3: the longest tumor diameter ≥ 5 cm; x 4: Barcelona staging B-C; x 5: microvascular invasion; x 6: moderate to low differentiation; x 7: incomplete capsule; x 8: nonanatomic hepatectomy). The Meta-analysis included 1 757 HCC cases, with 960 postoperative early recurrences and an early recurrence rate of 45.36%, finally the β 0 value was -0.201. The predictive model for 2-year recurrence of HCC was constructed and calculated as logit (P) = -0.201+0.835x 1+0.905x 2+0.783x 3+1.008x 4+0.765x 5+0.831x 6+1.533x 7+0.940x 8. Analysis of variance by external validation data showed that the differences in ascites, alpha-fetoprotein, tumor number, tumor diameter, Barcelona staging, microvascular invasion, tumor differentiation degree, capsule invasion, resection type, and systemic inflammation index were statistically significant between early recurrence group and early non-recurrence group (all P < 0.05). ROC curve analysis showed that AUC of postoperative early recurrence of HCC predicted by the model was 0.718, (95% CI: 0.689-0.753), the optimal cut-off value was 3.11, the Yoden index was 0.288, the sensitivity was 69.32%, and the specificity was 69.56%. Conclusions:The evidence-based prediction model constructed based on Meta-analysis for postoperative early recurrence of HCC has a high predictive value. However, further verification and optimization with big data is still needed.

Objective:To investigate the effect of theta-transcranial alternating current stimulation(θ-tACS)over the frontal-parietal cortex on working memory(WM)performance under different postural conditions(sitting and stand-ing)in healthy adults. Method:The study was a randomized,crossover,double-blind design.Sixteen healthy adults received tACS or sham stimulation in random order,with at least 72 hours interval between two sessions.Seven electrodes were placed in the left frontoparietal critical brain region with 6Hz for 21 mins.The sham stimulation used the same stimulation time and frequency as tACS group,but only 1 minute of stimulation(30s at the beginning and 30s at the end of the stimulation).WM performance was assessed before and immediately after stimulation,in-cluding accuracy,reaction time and inverse efficiency scores.Wilcoxon rank sum tests were used to observe between-and within-group differences on WM performance. Result:Fifteen participants completed the experiment and there were no significant differences in side effects be-tween two groups.There was a significant improvement in reaction time and inverse efficiency scores for sit-ting WM as well as reaction time for standing WM after the frontoparietal θ-tACS(P＜0.05).There was a trend towards improvement in inverse efficiency scores for standing WM(P=0.057),while there was no signif-icant difference in accuracy,reaction time and inverse efficiency scores of WM during the sitting and standing conditions in the sham stimulation(all P＞0.05). Conclusion:Frontoparietal θ-tACS can improve WM performance in both sitting and standing postures.This helps healthy adults cope with real-life multitasking cognitive activities and has important practical application value.

Objective : To examine the association between acute exposure to traffic-related air pollutants (TRAP) NOX and NO2 and outpatient visits of pediatric respiratory diseases. Methods : Data regarding outpatient visits to Department of Respiratory Diseases of Beijing Children's Hospital from 2015 to 2020 were collected, and the concentrations of nitrogen oxides (NOX), nitrogen dioxide (NO2) and other TRAP were collected from the surveillance sites assigned by the Peking University Health Science Center. A time-stratified case-crossover design was employed, and a conditional logistic regression model was created to examine the association between NOX and NO2 acute exposure and outpatient visits of pediatric respiratory diseases. Results : The daily mean outpatient visits of pediatric respiratory diseases were 571 (interquartile range, 554) person-times among children at ages of 0 to 14 years in Beijing Children's Hospital from 2015 to 2020, and the daily mean outpatient visits for upper respiratory tract infections (URI), bronchitis, and pneumonia were 265 (interquartile range, 282), 143 (interquartile range, 178) and 128 (interquartile range, 120) person-times, respectively. The daily mean concentrations of atmospheric NOX and NO2 were 67.8 (interquartile range, 50.7) and 49.3 (interquartile range, 30.7) μg/m3, respectively. Conditional logistic regression analysis showed the largest lagged effect of NOX and NO2 on pediatric respiratory diseases at cumulative lags of 0 to 7 days. An increase in NOX concentrations by an interquartile range resulted in the excess risks of URI, bronchitis and pneumonia by 6.87% (95%CI: 6.37%-7.38%), 7.25% (95%CI: 6.51%-7.99%), and 5.51% (95%CI: 4.69%-6.33%), and an increase in NO2 concentrations by an interquartile range resulted in excess risks of URI, bronchitis and pneumonia by 5.71% (95%CI: 5.12%-6.31%), 5.32% (95%CI: 4.51%-6.14%), and 4.83% (95%CI: 3.91%-5.75%), respectively. NOX and NO2 presented a more remarkable effect on outpatient visits of pediatric respiratory diseases among children at ages of over 5 years. Conclusion NOx and NO2 acute exposure may increase the outpatient visits of pediatric respiratory diseases.

BackgroundElderly cataract patients with depressive symptoms have heavier self-perceived burden， and worse vision-related quality of life compared to patients without depressive symptoms， resulting in the increase of family burden. Most previous studies showed that family relationship and vision were the main factors leading to depression in elderly cataract patients， and ignored the effects of self-feeling and complicated diseases on the psychological state of elderly cataract patients. ObjectiveTo explore the relationship between depressive symptoms， self-perceived burden and postoperative vision-related quality of life in elderly cataract patients， and to analyze the risk factors of depressive symptoms. So as to provide references for targeted psychological intervention in this population. MethodsA total of 104 elderly cataract patients admitted to Jiangsu Province Hospital （The First Affiliated Hospital with Nanjing Medical University） from July 1， 2020 to December 31， 2022 were included. General information， self-perceived burden and postoperative vision-related quality of life were investigated by self-designed questionnaire， Patients' Health Questionnaire depressive Scale-9 item （PHQ-9）， Self-Perceived Burden Scale （SPBS） and National Eye Institute Visual Function Questionnaire-25 （NEI-VFQ-25）. Pearson correlation analysis was performed to analyze the correlation among PHQ-9， SPBS and NEI-VFQ-25 scores. Logistic regression analysis was used to analyze the influencing factors of depressive symptoms in eldery cataract patients. ResultsA total of 100 elderly cataract patients completed an effective questionnaire survey. Among the participants， 31 cases （31.00%） were found to have depressive symptoms. The depressive group exhibited significantly higher SPBS score （t=11.062， P<0.01） and significantly lower NEI-VFQ-25 score （t=-5.235， P<0.01） than those of the non-depressive group. Pearson correlation analysis showed a positive correlation between PHQ-9 and SPBS score （r=0.485， P<0.01）， and a negative correlation between PHQ-9 and NEI-VFQ-25 score （r=-0.440， P<0.01）. Complicated with diabetes （OR=1.441， P<0.01）， osteoarthritis （OR=1.324， P<0.05） and high SPBS score （OR=1.340， P<0.05） were risk factors of depressive sympptoms in elderly cataract patients. ConclusionThe detection rate of depressive symptoms in elderly cataract patients is higher， and postoperative vision-related quality of life in elderly cataract patients complicated with depressive symptoms is poor. Risk factors include diabetes， osteoarthritis and high self-perceived burden. ［Funded by 2022 Jiangsu Province Elderly Health Scientific Research Project （number， LKM2022019）］

Humans ; Nitric Oxide ; Carcinoma, Non-Small-Cell Lung/drug therapy* ; Lung Neoplasms ; Lung ; Nitric Oxide Synthase ; Macrophages, Alveolar ; Pulmonary Alveoli