目的探讨AKI risk评分(基质金属蛋白酶-2×胰岛素样生长因子-7,TIMP-2×IGFBP-7)对急诊脓毒症患者死亡风险的预测价值。方法前瞻性观察2021年9月至2022年12月中国科学技术大学附属第一医院及北京协和医院急诊科收住的脓毒症患者...目的探讨AKI risk评分(基质金属蛋白酶-2×胰岛素样生长因子-7,TIMP-2×IGFBP-7)对急诊脓毒症患者死亡风险的预测价值。方法前瞻性观察2021年9月至2022年12月中国科学技术大学附属第一医院及北京协和医院急诊科收住的脓毒症患者,分别测量患者入院时和入院后6 h的AKI risk评分并计算其变化值(AKI risk-gap),利用多因素Logistic回归、Cox回归、受试者工作特征(ROC)曲线及曲线下面积(AUC)分析AKI risk评分对患者院内死亡风险的预测效能;亚组分析中根据患者是否罹患AKI进一步分析AKI risk评分与不同亚组(AKI组和非AKI组)患者预后的关系。结果本研究共纳入患者202例,住院期间死亡87例(43%)。ROC曲线显示,6 h AKI risk评分预测脓毒症患者院内死亡最为准确,其AUC为0.71(95%CI 0.63~0.78)。亚组分析中AKI组患者6 h AKI risk评分预测院内死亡的AUC为0.76(95%CI 0.65~0.85),非AKI组AUC为0.63(95%CI 0.52~0.73)。多因素Logistic回归和Cox回归分析表明,6 h AKI risk评分和AKI risk-gap是患者院内死亡的独立危险因素。结论AKI risk评分对脓毒症患者院内死亡风险有较好的预测价值,尤其6 h AKI risk评分在罹患AKI的亚组患者中预测价值最高,可为临床区分高危患者并给予相应治疗提供参考。展开更多
基于目前AVO(Amplitude Versus Offset)反演技术中所普遍使用的Aki-Richards和Shuey近似,以Zoeppritz方程的准确解为参照,针对几种典型油藏反射界面,计算了Aki-Richards和Shuey近似所产生的相对误差,绘出了相对误差随入射角的变化曲线....基于目前AVO(Amplitude Versus Offset)反演技术中所普遍使用的Aki-Richards和Shuey近似,以Zoeppritz方程的准确解为参照,针对几种典型油藏反射界面,计算了Aki-Richards和Shuey近似所产生的相对误差,绘出了相对误差随入射角的变化曲线.计算表明,在存在极性反转的反射界面,Aki-Richards和Shuey近似的计算结果会带来很大的误差,在极性反转点附近的计算结果失去它的可信性;在泥岩与气砂、油砂或水砂组成的反射界面,即便就是在小角度范围内也存在较大的计算误差(如泥岩—气砂反射界,入射角在5°之内,相对误差就超过了10%);在油砂和水砂反射界面Aki Richards和Shuey是很好的近似.本文对Aki-Richards和Shuey近似结合几种典型油藏反射界面给出定量的分析,为储层参数反演方案的设计特别是反演道集及范围的选择提供理论依据.展开更多
为了及早发现重症监护室中的急性肾损伤高危患者,为其提供适当的护理,实现医疗资源的合理利用,研究建立因果贝叶斯网络模型进行急性肾损伤高危患者死亡风险预测。从重症监护医学信息市场(Medical Information Mart for Intensive CareⅢ...为了及早发现重症监护室中的急性肾损伤高危患者,为其提供适当的护理,实现医疗资源的合理利用,研究建立因果贝叶斯网络模型进行急性肾损伤高危患者死亡风险预测。从重症监护医学信息市场(Medical Information Mart for Intensive CareⅢ,MIMIC-Ⅲ)数据库中筛选了25个研究变量和3870条患者数据,使用因果发现算法进行特征降维。通过NO TEARS算法构建因果图并建立因果贝叶斯网络进行实验,通过机器学习算法验证重要特征的合理性,并对网络结构进行因果效应估计,模型具有最高的受试者工作特征曲线下面积(Area Under the Receiver Operating Characteristic,AUROC)分数,为81.7%,优于逻辑回归(Logistic Regression,LR)、随机森林(Random Forest,RF)和极端梯度提升树(eXtreme Gradient Boosting,XGBoost)。此外,模型的重要特征预测能力在各种建模中都很稳健,构建的因果贝叶斯网络具有更好的预测效果并具备良好的解释能力。展开更多
Context: Acute kidney injury (AKI) in intensive care unit (ICU) is common and associated with very high mortality. In Togo, a tropical country with limited resources and only one nephrology department in the north, ac...Context: Acute kidney injury (AKI) in intensive care unit (ICU) is common and associated with very high mortality. In Togo, a tropical country with limited resources and only one nephrology department in the north, acute kidney injury seems to be a real tragedy with high mortality. Aims: to determine risk factors for mortality in acute kidney injury in the intensive care units. Methods and Material: We made a multicentric cross sectional study during 6 months in the four referral centers in northern Togo. Univariate and multivariate logistic regression was used to identify factors associated with mortality. Data were analyzed using RStudio 2023.04.1. Results: A total of 12.6% of patients admitted to intensive care had presented with AKI. The mean age was 49.6 ± 17.9. The sex ratio (M/F) was 2.1. Community-acquired AKI was in the majority (67.7%). Oligo anuria was the most frequent functional sign (38.4%). In our series, 81.6% of patients were in KDIGO stages 2 to 3. AKI was organic in 56.2% of cases. Mortality was 44.3%. In multivariate analysis, the main factors predictive of death were: respiratory distress (OR = 2.36;p Conclusions: Acute kidney injury in intensive care is common in northern Togo, and mortality is high. Identification of associated factors should help anticipate prognosis.展开更多
文摘目的探讨AKI risk评分(基质金属蛋白酶-2×胰岛素样生长因子-7,TIMP-2×IGFBP-7)对急诊脓毒症患者死亡风险的预测价值。方法前瞻性观察2021年9月至2022年12月中国科学技术大学附属第一医院及北京协和医院急诊科收住的脓毒症患者,分别测量患者入院时和入院后6 h的AKI risk评分并计算其变化值(AKI risk-gap),利用多因素Logistic回归、Cox回归、受试者工作特征(ROC)曲线及曲线下面积(AUC)分析AKI risk评分对患者院内死亡风险的预测效能;亚组分析中根据患者是否罹患AKI进一步分析AKI risk评分与不同亚组(AKI组和非AKI组)患者预后的关系。结果本研究共纳入患者202例,住院期间死亡87例(43%)。ROC曲线显示,6 h AKI risk评分预测脓毒症患者院内死亡最为准确,其AUC为0.71(95%CI 0.63~0.78)。亚组分析中AKI组患者6 h AKI risk评分预测院内死亡的AUC为0.76(95%CI 0.65~0.85),非AKI组AUC为0.63(95%CI 0.52~0.73)。多因素Logistic回归和Cox回归分析表明,6 h AKI risk评分和AKI risk-gap是患者院内死亡的独立危险因素。结论AKI risk评分对脓毒症患者院内死亡风险有较好的预测价值,尤其6 h AKI risk评分在罹患AKI的亚组患者中预测价值最高,可为临床区分高危患者并给予相应治疗提供参考。
文摘基于目前AVO(Amplitude Versus Offset)反演技术中所普遍使用的Aki-Richards和Shuey近似,以Zoeppritz方程的准确解为参照,针对几种典型油藏反射界面,计算了Aki-Richards和Shuey近似所产生的相对误差,绘出了相对误差随入射角的变化曲线.计算表明,在存在极性反转的反射界面,Aki-Richards和Shuey近似的计算结果会带来很大的误差,在极性反转点附近的计算结果失去它的可信性;在泥岩与气砂、油砂或水砂组成的反射界面,即便就是在小角度范围内也存在较大的计算误差(如泥岩—气砂反射界,入射角在5°之内,相对误差就超过了10%);在油砂和水砂反射界面Aki Richards和Shuey是很好的近似.本文对Aki-Richards和Shuey近似结合几种典型油藏反射界面给出定量的分析,为储层参数反演方案的设计特别是反演道集及范围的选择提供理论依据.
文摘为了及早发现重症监护室中的急性肾损伤高危患者,为其提供适当的护理,实现医疗资源的合理利用,研究建立因果贝叶斯网络模型进行急性肾损伤高危患者死亡风险预测。从重症监护医学信息市场(Medical Information Mart for Intensive CareⅢ,MIMIC-Ⅲ)数据库中筛选了25个研究变量和3870条患者数据,使用因果发现算法进行特征降维。通过NO TEARS算法构建因果图并建立因果贝叶斯网络进行实验,通过机器学习算法验证重要特征的合理性,并对网络结构进行因果效应估计,模型具有最高的受试者工作特征曲线下面积(Area Under the Receiver Operating Characteristic,AUROC)分数,为81.7%,优于逻辑回归(Logistic Regression,LR)、随机森林(Random Forest,RF)和极端梯度提升树(eXtreme Gradient Boosting,XGBoost)。此外,模型的重要特征预测能力在各种建模中都很稳健,构建的因果贝叶斯网络具有更好的预测效果并具备良好的解释能力。
文摘Context: Acute kidney injury (AKI) in intensive care unit (ICU) is common and associated with very high mortality. In Togo, a tropical country with limited resources and only one nephrology department in the north, acute kidney injury seems to be a real tragedy with high mortality. Aims: to determine risk factors for mortality in acute kidney injury in the intensive care units. Methods and Material: We made a multicentric cross sectional study during 6 months in the four referral centers in northern Togo. Univariate and multivariate logistic regression was used to identify factors associated with mortality. Data were analyzed using RStudio 2023.04.1. Results: A total of 12.6% of patients admitted to intensive care had presented with AKI. The mean age was 49.6 ± 17.9. The sex ratio (M/F) was 2.1. Community-acquired AKI was in the majority (67.7%). Oligo anuria was the most frequent functional sign (38.4%). In our series, 81.6% of patients were in KDIGO stages 2 to 3. AKI was organic in 56.2% of cases. Mortality was 44.3%. In multivariate analysis, the main factors predictive of death were: respiratory distress (OR = 2.36;p Conclusions: Acute kidney injury in intensive care is common in northern Togo, and mortality is high. Identification of associated factors should help anticipate prognosis.
