目的基于临床病历数据构建并验证急性缺血性卒中取栓后继发出血转化风险的列线图模型。方法通过检索病历系统获取2015-01-2022-08广东省中医院神经科介入取栓治疗急性缺血卒中的临床资料,以7∶3比例将筛选所得病例随机分为训练集及验证...目的基于临床病历数据构建并验证急性缺血性卒中取栓后继发出血转化风险的列线图模型。方法通过检索病历系统获取2015-01-2022-08广东省中医院神经科介入取栓治疗急性缺血卒中的临床资料,以7∶3比例将筛选所得病例随机分为训练集及验证集。运用R4.1.3软件进行Lasso回归及多因素Logistic回归分析确定影响取栓治疗后继发出血转化的独立预后因素,通过一致性指数(C-index)、受试者工作特征曲线(receiver operator characteristic,ROC)及校准曲线评估列线图的可靠性。结果共检索录入382例患者临床资料,训练集270例,验证集112例,经Lasso回归及多因素Logistic回归后,确定年龄、术前收缩压、Alberta卒中项目早期CT评分(Alberta Stroke Program Early CT Score,ASPECTS)、心源性栓塞、手术时长及再灌注损伤为继发出血转化的独立预后因素。训练集与验证集的C-index分别为0.846(95%CI:0.780~0.912)、0.845(95%CI:0.759,0.931),训练集与验证集AUC值分别为0.846、0.845。模型的校正曲线显示预测值与实际观测值具有良好一致性,决策曲线分析(decision curve analysis,DCA)显示列线图具有较好的临床获益性。结论年龄、术前收缩压、ASPECTS、心源性栓塞、手术时长及再灌注损伤是急性前循环缺血性卒中行介入取栓后继发出血转化的独立危险因素,本研究构建的列线图模型能较准确的预测急性缺血卒中患者介入取栓后继发出血转化的风险,其广泛应用仍需进行外部和前瞻性验证。展开更多
Pneumatic proportional control servo regulator is the core component of a pneumatic-loading experimental system,which is very important in solving the overcharging problem.However,previous research on control of pneum...Pneumatic proportional control servo regulator is the core component of a pneumatic-loading experimental system,which is very important in solving the overcharging problem.However,previous research on control of pneumatic proportional regulator in a pneumatic-loading experimental system led to failure in analysis of the influence of opening error of the switch regulator because it did not analyze the regulator basic working principle and process.The traditional control method cannot fully solve the overcharging problem nor ensure adequate control performance of the regulator.After seriously studying the working principle and key mechanical parameters of the valve,a fuzzy parameter-adaptive controller is designed by introducing a linear mixture of the pressure and opening errors of the switch regulator to reduce pressure overshoot and optimize its control performance.According to the fuzzy-control strategy based on the working characteristics and mechanical parameters of the valve,the overshoot phenomenon of the pneumatic-loading system is solved,and the pressure overshoot is eliminated.The error of the output air pressure of the regulator is 1.24%,which is small.The adjustable pressure range of the regulator is 0.2–0.6 MPa.The maximum deviation is 0.012 MPa.The linearity of the case is 1.34%F.S.展开更多
文摘目的基于临床病历数据构建并验证急性缺血性卒中取栓后继发出血转化风险的列线图模型。方法通过检索病历系统获取2015-01-2022-08广东省中医院神经科介入取栓治疗急性缺血卒中的临床资料,以7∶3比例将筛选所得病例随机分为训练集及验证集。运用R4.1.3软件进行Lasso回归及多因素Logistic回归分析确定影响取栓治疗后继发出血转化的独立预后因素,通过一致性指数(C-index)、受试者工作特征曲线(receiver operator characteristic,ROC)及校准曲线评估列线图的可靠性。结果共检索录入382例患者临床资料,训练集270例,验证集112例,经Lasso回归及多因素Logistic回归后,确定年龄、术前收缩压、Alberta卒中项目早期CT评分(Alberta Stroke Program Early CT Score,ASPECTS)、心源性栓塞、手术时长及再灌注损伤为继发出血转化的独立预后因素。训练集与验证集的C-index分别为0.846(95%CI:0.780~0.912)、0.845(95%CI:0.759,0.931),训练集与验证集AUC值分别为0.846、0.845。模型的校正曲线显示预测值与实际观测值具有良好一致性,决策曲线分析(decision curve analysis,DCA)显示列线图具有较好的临床获益性。结论年龄、术前收缩压、ASPECTS、心源性栓塞、手术时长及再灌注损伤是急性前循环缺血性卒中行介入取栓后继发出血转化的独立危险因素,本研究构建的列线图模型能较准确的预测急性缺血卒中患者介入取栓后继发出血转化的风险,其广泛应用仍需进行外部和前瞻性验证。
基金supported by the China Postdoctoral Science Foundation(Grant No. 2019M660392)the Open Research Project of the State Key Laboratory of Media Convergence and Communication, Communication University of China (Grant No. SKLMCC2020KF002)the National Key Research and Development Project (Grant Nos. 2019YFC0121700,2021YFC0122502)
文摘Pneumatic proportional control servo regulator is the core component of a pneumatic-loading experimental system,which is very important in solving the overcharging problem.However,previous research on control of pneumatic proportional regulator in a pneumatic-loading experimental system led to failure in analysis of the influence of opening error of the switch regulator because it did not analyze the regulator basic working principle and process.The traditional control method cannot fully solve the overcharging problem nor ensure adequate control performance of the regulator.After seriously studying the working principle and key mechanical parameters of the valve,a fuzzy parameter-adaptive controller is designed by introducing a linear mixture of the pressure and opening errors of the switch regulator to reduce pressure overshoot and optimize its control performance.According to the fuzzy-control strategy based on the working characteristics and mechanical parameters of the valve,the overshoot phenomenon of the pneumatic-loading system is solved,and the pressure overshoot is eliminated.The error of the output air pressure of the regulator is 1.24%,which is small.The adjustable pressure range of the regulator is 0.2–0.6 MPa.The maximum deviation is 0.012 MPa.The linearity of the case is 1.34%F.S.
