Objective:To construct a network calculator based on interpretable machine learning models to predict the risk of prolonged length of stay in patients with acute ischemic stroke (AIS), and to provide a tool for the development of individualized intervention plans for patients.Methods:Adopting a retrospective analysis method. The 537 patients with AIS admitted to the Nanyang Central Hospital from July 2022 to July 2024 were selected. Length of stay exceeding the median length of stay was defined as length of stay prolongation. Length of stay prolongation risk profile variables were screened by Boruta algorithm. The 537 patients were randomly divided into a training set (322 cases) and a test set (215 cases) in a 3:2 ratio according to the random number table method to construct and train nine machine learning models and perform tenfold cross-validation. The best predictive performance model was assessed using receiver operating characteristic (ROC) curve analysis and calculating the area under the curve (AUC). The predictive accuracy and clinical utility of the model was assessed using calibration curve, clinical impact curve and decision curve analyses. Additional interpretation and visualisation of the machine learning model using Shapley additive explanations (SHAP) bar charts, summary, dependency and force diagrams. A network calculator for predicting the risk of length of stay prolongation in patients with AIS was constructed using the corresponding R package.Results:Among 537 AIS patients, there were 312 males and 225 females with an age of (58.40 ± 9.00) years old. The incidence of length of stay prolongation was 43.0% (231/537). Boruta algorithm screened 10 characteristic variables. The results of ROC curve analysis showed that the AUC of the extreme gradient boosting (XGBoost) model was at the highest among the 9 machine learning models in 10 random samples. In the training and test sets, the calibration curve and clinical impact curve analysis showed that the C-index was 0.815 and 0.816, respectively, indicating high consistency between the predicted results of the XGBoost model and the actual observations. Decision curve analysis showed that the net clinical benefit was>0 when the risk threshold was>0.18 and >0.22, respectively, indicating that the model had a high application value in actual clinical decision-making. The SHAP bar graph showed the order of importance as pneumonia, urinary tract infection, age, myoglobin, triacylglycerol, neuron-specific enolase (NSE), hemoglobin, total cholesterol, homocysteine (HCY), and National Institute of Health Stroke Scale (NIHSS) scores. SHAP summary charts visualised the contribution of the 10 characteristic variables, which showed a 'bipolar distribution' phenomenon. SHAP dependency plots showed the dependency between the observed values of the 10 variables and the SHAP values, with the trend being most significant for patients with pneumonia. SHAP seeks to provide a local interpretation for individual samples, making the XGBoost model more transparent and interpretable. A web-based calculator (https://nomogram1203.shinyapps.io/LOS_web/) based on the interpretable XGBoost model dynamically quantifies the risk of prolonged length of stay in patients with AIS.Conclusions:Pneumonia, urinary tract infection, age, myoglobin, hemoglobin, total cholesterol, NSE, HCY, triacylglycerol, and NIHSS can effect length of stay prolongation in AIS patients, and a network calculator constructed based on the interpretable XGBoost model dynamically predicts the risk of length of stay prolongation in patients with AIS, which can help to achieve an accurate risk assessment for individual patients.

Objective:To construct a network calculator based on interpretable machine learning models to predict the risk of prolonged length of stay in patients with acute ischemic stroke (AIS), and to provide a tool for the development of individualized intervention plans for patients.Methods:Adopting a retrospective analysis method. The 537 patients with AIS admitted to the Nanyang Central Hospital from July 2022 to July 2024 were selected. Length of stay exceeding the median length of stay was defined as length of stay prolongation. Length of stay prolongation risk profile variables were screened by Boruta algorithm. The 537 patients were randomly divided into a training set (322 cases) and a test set (215 cases) in a 3:2 ratio according to the random number table method to construct and train nine machine learning models and perform tenfold cross-validation. The best predictive performance model was assessed using receiver operating characteristic (ROC) curve analysis and calculating the area under the curve (AUC). The predictive accuracy and clinical utility of the model was assessed using calibration curve, clinical impact curve and decision curve analyses. Additional interpretation and visualisation of the machine learning model using Shapley additive explanations (SHAP) bar charts, summary, dependency and force diagrams. A network calculator for predicting the risk of length of stay prolongation in patients with AIS was constructed using the corresponding R package.Results:Among 537 AIS patients, there were 312 males and 225 females with an age of (58.40 ± 9.00) years old. The incidence of length of stay prolongation was 43.0% (231/537). Boruta algorithm screened 10 characteristic variables. The results of ROC curve analysis showed that the AUC of the extreme gradient boosting (XGBoost) model was at the highest among the 9 machine learning models in 10 random samples. In the training and test sets, the calibration curve and clinical impact curve analysis showed that the C-index was 0.815 and 0.816, respectively, indicating high consistency between the predicted results of the XGBoost model and the actual observations. Decision curve analysis showed that the net clinical benefit was>0 when the risk threshold was>0.18 and >0.22, respectively, indicating that the model had a high application value in actual clinical decision-making. The SHAP bar graph showed the order of importance as pneumonia, urinary tract infection, age, myoglobin, triacylglycerol, neuron-specific enolase (NSE), hemoglobin, total cholesterol, homocysteine (HCY), and National Institute of Health Stroke Scale (NIHSS) scores. SHAP summary charts visualised the contribution of the 10 characteristic variables, which showed a 'bipolar distribution' phenomenon. SHAP dependency plots showed the dependency between the observed values of the 10 variables and the SHAP values, with the trend being most significant for patients with pneumonia. SHAP seeks to provide a local interpretation for individual samples, making the XGBoost model more transparent and interpretable. A web-based calculator (https://nomogram1203.shinyapps.io/LOS_web/) based on the interpretable XGBoost model dynamically quantifies the risk of prolonged length of stay in patients with AIS.Conclusions:Pneumonia, urinary tract infection, age, myoglobin, hemoglobin, total cholesterol, NSE, HCY, triacylglycerol, and NIHSS can effect length of stay prolongation in AIS patients, and a network calculator constructed based on the interpretable XGBoost model dynamically predicts the risk of length of stay prolongation in patients with AIS, which can help to achieve an accurate risk assessment for individual patients.

Objective:To investigate the molecular mechanism of antiglioma effect of curcumin.Methods:Cell experiment: (1) U251MG and SHG-44 cells at logarithmic growth phase were treated with 10 μmol/L curcumin (curcumin group) or same volume of dimethyl sulfoxide solution (control group); cells were transfected with negative control small interfering RNA (siRNA) and long non-coding RNA (lncRNA) H19 siRNA (negative control siRNA group and H19 siRNA group); cells were transfected with negative control siRNA and H19 siRNA, respectively, and then, they were treated with 10 μmol/L curcumin (negative control siRNA+curcumin group and H19 siRNA+curcumin group); the H19 siRNA was co-transfected with negative control miR inhibitor or miR-491-5p inhibitor into these cells (H19 siRNA+negative control inhibitor group and H19 siRNA+miR-4915p inhibitor group); H19 siRNA+negative control miR inhibitor or H19 siRNA+miR-491-5p inhibitor were co-transfected into the cells, and then, they were treated with 10 μmol/L curcumin (H19 siRNA+negative control inhibitor+curcumin group and H19 siRNA+miR-491-5p inhibitor+curcumin group); the cells were co-transfected with miR-491-5p mimic+blank plasmid or miR-491-5p mimic+HOXA9 overexpression plasmid, and then they were treated with 10 μmol/L curcumin (miR-491-5p mimic+blank plasmid+curcumin group and miR-491-5p mimic+HOXA9 overexpression plasmid+curcumin group); real-time fluorescent quantitative PCR (qRT-PCR) was used to detect the mRNA expressions of H19, miR-491-5p, and HOXA9; CCK-8 assay was used to detect the cell proliferation; flow cytometry was used to detect the cell apoptosis; plate cloning method was employed to detect the number of cell clone formation; Transwell assay was used to detect the cell migration; and the HOXA9 protein expression was measured by Western blotting. (2) The 293T cells at the logarithmic growth phase were chosen; the negative control miRNA mimics or miR-491-5p mimics combined with wild-type H19, wild-type HOXA9 3'-UTR plasmid vectors were co-transfected into the cells, respectively (negative control mimic+wild type H19 group and miR-491-5p mimic+wild type H19 group, negative control mimic+wild type HOXA9 3'-UTR group and miR-491-5p mimic+wild type HOXA9 3'-UTR group); the luciferase activity was detected by dual luciferase reporter experiment. (3) Thirty specimens from glioma patients (glioma group) underwent surgical resection and pathologically confirmed in our hospital from May 2017 to May 2019 and 30 normal brain tissue specimens obtained during decompression (normal group) at the same period were chosen; the mRNA expressions of H19, miR-491-5p, and HOXA9 were detected by qRT-PCR, and the HOXA9 protein expression level in these specimens was detected by Western blotting. (4) Twenty-four nude mice were randomly divided into negative control short hairpin RNA (shRNA) group, H19 shRNA group, negative control shRNA+curcumin group, and H19 shRNA+curcumin group ( n=6); U251MG cells stably transfected with negative control shRNA or H19 shRNA were intraperitoneally injected, respectively, into the mice; and 60 mg/kg curcumin was injected on the next d; the tumor volume was measured on the 7 th, 11 th, 15 th, 19 th, 23 rd, and 27 th d of rearing; and the H19, miR-491-5p and HOXA9 mRNA expressions in the tumor tissues were detected by qRT-PCR; the HOXA9 protein expression was detected by Western blotting. Results:(1) When curcumin group comparing with control group, and H19 siRNA group comparing with negative control siRNA group, U251MG and SHG-44 cells had significantly decreased miR-491-5p mRNA and protein expressions, and significantly increased miR-491-5p mRNA expression ( P<0.05); as compared with that in the H19 siRNA+negative control inhibitor group, the HOXA9 mRNA and protein expressions in U251MG and SHG-44 cells of H19 siRNA+miR-491-5p inhibitor group were significantly higher ( P<0.05). When curcumin group comparing with control group, H19 siRNA group comparing with negative control siRNA group, H19 siRNA+curcumin group comparing with negative control siRNA+curcumin group, the U251MG and SHG-44 cells after 72 h of culture had significantly decreased cell proliferation rate, significantly increased apoptosis rate, significantly reduced number of cell clone formation, and significantly reduced cell migration number ( P<0.05). When H19 siRNA+miR-491-5p inhibitor+curcumin group comparing with H19 siRNA+negative control inhibitor+curcumin group, miR-491-5p mimic+HOXA9 overexpression plasmid+curcumin group comparing with miR-491-5p mimic+blank plasmid+curcumin group, the U251MG and SHG-44 cells after 72 h of culture had significantly increased cell proliferation rate, significantly reduced apoptosis rate, significantly increased number of cell clone formation, and significantly increased cell migration number ( P<0.05). (2) When miR-491-5p mimic+wild-type H19 group comparing with negative control mimic+wild-type H19 group, miR-491-5p mimic+wild-type HOXA9 3'-UTR group comparing with negative control mimic+wild-type HOXA9 3'-UTR group, the cell luciferase activity was significantly reduced ( P<0.05). (3) As compared with those in the normal group, the H19 and HOXA9 mRNA expressions and HOXA9 protein expression in the glioma group were significantly increased, and the miR-491-5p mRNA expression was significantly reduced ( P<0.05). (4) On the 27 th d of rearing, when H19 shRNA group comparing with negative control shRNA group, and H19 shRNA+curcumin group comparing with negative control shRNA+curcumin group, the tumor volume was significantly reduced, the miR-491-5p mRNA expression in the tumor tissues was significantly increased, and the H19 mRNA, HOXA9 mRNA and protein expressions were significantly reduced ( P<0.05). Conclusion:Curcumin may inhibit the cell proliferation and migration and promote the apoptosis of glioma cells through lncRNA H19/miR-491-5p/HOXA9 axis.

Objective To observe the effects of different concentrations of baicalin on the mouse model of experimental autoimmune encephalomyelitis (EAE) and to explore its mechanisms.Methods A mouse model of EAE was established with MOG33-55 peptide and bacillus Calmette-Guerin (BGG) vaccine with complete Freund adjuvant (CFA).At the third day after immunization, high and low doses of baicalin were administered to the mice intragastrically once a day for 20 days.The neurological function of mice was evaluated.TUNEL staining was used to detect apoptosis in the spinal cord tissue.The level of ATP in spinal cord tissue was detected by an ATP determination kit.Furthermore, the protein expressions of Bax, Bcl-2, cleaved cas-3 and cleaved cas-9 were detected by western blot, respectively.Results Baicalin improved the neurological function and delayed the onset time in EAE mice.