Objective:To establish a machine learning-based risk prediction model for perimenopausal depressive symptoms and to identify associated risk factors.Methods:A total of 1 105 women aged 45 to 55 years were selected from the 2020 China Health and Retirement Longitudinal Study (CHARLS) dataset.Three machine learning algorithms, including Random Forest, XGBoost and Adaptive Boosting (AdaBoost), were employed to construct prediction models for perimenopausal depressive symptoms. Descriptive statistics and between-group comparisons were performed using SPSS 24.0.And Python 3.10 software was used to build the risk prediction model. Model performance was assessed using receiver operating characteristic (ROC) curves and calibration plots, and the optimal model was identified accordingly. The Shapley additive explanation (SHAP) algorithm was then used to analyze feature importance and the influence of each predictor on the outcome.Results:Among the 1 105 perimenopausal women, 671(60.7%)were categorized in the non-depressive group and 434 (39.3%) in the depressive group. The Random Forest model demonstrated the best overall predictive performance among the three machine learning models, achieving an area under the ROC curve (AUC) of 0.793 and a calibration error of 0.181. SHAP analysis revealed that annual household income was the strongest risk factor in the Random Forest model, with a relative importance of 0.048, followed by cognitive function(0.047), self-rated health status(0.046), life satisfaction(0.043), sleep duration(0.041).Conclusions:The Random Forest based model effectively predicts the risk of perimenopausal depressive symptoms. Annual household income, cognitive function, self-rated health, and life satisfaction are risk factors for depressive symptoms in perimenopausal women.

Objective:To establish a machine learning-based risk prediction model for perimenopausal depressive symptoms and to identify associated risk factors.Methods:A total of 1 105 women aged 45 to 55 years were selected from the 2020 China Health and Retirement Longitudinal Study (CHARLS) dataset.Three machine learning algorithms, including Random Forest, XGBoost and Adaptive Boosting (AdaBoost), were employed to construct prediction models for perimenopausal depressive symptoms. Descriptive statistics and between-group comparisons were performed using SPSS 24.0.And Python 3.10 software was used to build the risk prediction model. Model performance was assessed using receiver operating characteristic (ROC) curves and calibration plots, and the optimal model was identified accordingly. The Shapley additive explanation (SHAP) algorithm was then used to analyze feature importance and the influence of each predictor on the outcome.Results:Among the 1 105 perimenopausal women, 671(60.7%)were categorized in the non-depressive group and 434 (39.3%) in the depressive group. The Random Forest model demonstrated the best overall predictive performance among the three machine learning models, achieving an area under the ROC curve (AUC) of 0.793 and a calibration error of 0.181. SHAP analysis revealed that annual household income was the strongest risk factor in the Random Forest model, with a relative importance of 0.048, followed by cognitive function(0.047), self-rated health status(0.046), life satisfaction(0.043), sleep duration(0.041).Conclusions:The Random Forest based model effectively predicts the risk of perimenopausal depressive symptoms. Annual household income, cognitive function, self-rated health, and life satisfaction are risk factors for depressive symptoms in perimenopausal women.

Objective To observe the correlation between CEUS quantitative parameters of carotid plaques and leukocytes in patients with acute ischemic stroke.Methods Sixty-two patients with large artery atherosclerosis stroke (LAAS group) confirmed by CT or MRI were enrolled,while 54 patients in the same period of hospitalization,age and gender-matched,no history of cardiovascular events with atherosclerosis were taken as control group.The correlation between CEUS quantitative parameters of carotid plaques and leukocytes in two groups were compared.Multiple linear regression model was built and the risk factors of CEUS quantitative parameters were analyzed.Results The total leukocytes count,neutrophils count and neutrophil/lymphocyte ratio in LAAS group were higher,while the lymphocytes count was lower than those in control group (all P＜0.05).CEUS parameters,including timeqntensity curve (TIC) peak (TIC-P),mean (TIC-M),fitting curve (FC) peak (FC-P),sharpness (FC-S) and area under the curve (FC-AUC) of carotid plaques were higher than those in control group (all P＜0.05),while neutrophils count and neutrophil/lymphocyte ratio were positively correlated with FC-AUC (r=0.298 and 0.739,respectively;all P＜0.05).Total leukocytes count was independent risk factor of TIC-P,and neutrophil/lymphocyte ratio was independent risk factor of FC-AUC (all P＜0.05).Conclusion CEUS quantitative parameters of carotid plaques related to leukocytes count.Increased leukocytes or neutrophil/ lymphocyte ratio might rise vulnerability of plaques.

Objective To detecte the neovascularizations in carotid plaques using contrast-enhanced ultrasound (CEUS) and explore the relationship between the neovascularizations and the peripheral leukocytes in the patients with acute cerebral infarction. Methods Sixty-two patients with large artery atherosclerosis cerebral infarction were selected as cerebral infarction group;and 54 age-and gender-matched patients with atherosclerosis and without cerebral cerebrovascular events were recruited control group.The dominant carotid artery plaques were performed by CEUS,the peak of time-intensity curve(TIC-P) and the mean of time-intensity curve (TIC-M) were measured by off-line quantitative analysis.The peak (FC-P), time to peak (FC-TP),sharpness (FC-S) and under the curve area (FC-AUC) were obtained from fitting curves of time-intensity. The correlations between parameters of contrast-enhanced ultrasound and leukocyte counts were analyzed by Pearson correlation analysis. Results ①In the cerebral infarction group, the total leukocytes and neutrophils were higher than those in the control group,while the number of lymphocytes was lower than that of the control group(all P <0.05). ②In the cerebral infarction group,the TIC-P and TIC-M values were greater than those in the control group ( P < 0.05).Compared with the control group,the FC-P,FC-S and FC-AUC values in the cerebral infarction group were increased (all P <0.05). ③There was a negative correlation between PIG-P,TIC-M of FC-P and lymphocytes( r = -0.291,-0.263 and -0.270;all P <0.05).FC-S and FC-AUC were positively correlated with neutrophils ( r =0.261,0.298;all P < 0.05). Conclusions Carotid plaque neovascularizations is related to peripheral leukocyte count.CEUS help us know more the vulnerability of plaque.

Objective: To explore the molecular mechanism of miR-195-5p targeting FGF2 to inhibit the proliferation, apoptosis, invasion and migration of endometrial cancer HEC-1B cells. Methods: After culture and transfection, HEC-1B cells were divided into 4 groups: HEC-1B group, miR-195-5p mimic group, pLV-FGF2 group and miR-195-5p+FGF2 group. The expressions of miR-195-5p and mRNA levels of FGF2 were detected by qRT-PCR. The targeted relationship of miR-195-5p and FGF2 was verified by luciferase assay. The protein expression of FGF2 was examined by Western blotting; Proliferation of HEC-1B cells was measured by CCK-8; Apoptosis was tested by flow cytometry; HEC-1B cell invasion was detected by transwell, and migration was measured by scratch assay. Results: Compared with HEC-1B group, the expression of miR-195-5p in miR-195-5p mimic group was elevated while FGF2 mRNA level was declined (all P<0.01). Luciferase assay indicated that FGF2 was a target of miR-195-5p. Compared with HEC-1B group, the protein level of FGF2 in miR-195-5p mimic group was decreased, and the protein levels of FGF2 in pLV-FGF2 group were enhanced (P<0.01). The protein levels of FGF2 in miR-195-5p+FGF2 group were lower than that in pLV-FGF2 group (all P<0.01). The proliferation in miR-195-5p mimic group was lower than HEC-1B group (P<0.01), while the proliferation in pLV-FGF2 group was higher than that in HEC-1B group (all P<0.01). Compared with HEC-1B group, apoptosis in miR-195-5p mimic group was increased, and apoptosis in pLV-FGF2 group was decreased (P<0.01); moreover, apoptosis in miR-195-5p+FGF2 group was higher than that in pLV-FGF2 group (P<0.01). Compared with HEC-1B group, the number of invasive cells per field and the rate of wound healing in miR195-5p mimic group were decreased, while those in pLV-FGF2 group was enhanced (P<0.01); moreover, the number of invasive cells per field and the rate of wound healing in miR-195-5p+FGF2 group was lower than those in pLV-FGF2 group (all P<0.01). Conclusion: miR-195-5p inhibits proliferation, invasion and migration and promotes apoptosis of endometrial cancer HEC-1B cells by targeting FGF2, and could be used as a treatment target of endometrial cancer.

Objective To summarize the nursing experience of treatment of percutaneous and transhepatic portal venous autologous bone marrow stem cell transplantation for chronic hepatic failure.Methods 19 patients who were definitely diagnosed as chronic liver failure received pertinent nursing in different perioperative period of transhepatic portal venous autologous bone marrow stem cell transplantation.Results All of the 19 patients went through perioperative period safely without any adverse reactions or complications.Conclusions In the process of treatment of autologous bone marrow stem cell transplantation for chronic hepatic failure,sufficient preoperative preparation,good communication and close cooperation among doctors,nurses and patients during operation,careful nursing and rehabilitation instruction after operation,are important assurances for autologous bone marrow stem cell transplantation to run smoothly.