Evolution of Proteinuria and Renal Function in Women with Pre-Eclampsia at the Gynecology Department of the Teaching Hospital of Cocody

Background: Pre-eclampsia has long been considered as a disease that disappears after the removal of the placenta. It has now been shown that its symptoms can persist for months after giving birth. Objectives: To study the evolution of proteinuria and renal function in women with pre-eclampsia. Patients and Methods: An analytical prospective study was carried out in the Hospitalization Unit of the Gynecology Department of the Teaching Hospital of Cocody (Abidjan) from May 3, 2021 to November 15, 2021. It focused on the follow-up of proteinuria and renal function in 50 women who had pre-eclampsia during their pregnancy, in the three months following their delivery. Results: The average age of the patients was 30.38 ± 6 years (range 18 and 40 years). Thirty-two percent were nulliparous and 62% had no risk factors for pre-eclampsia. The diagnosis of pre-eclampsia was made in 52% of cases before 37 weeks of amenorrhea. Sixty-two percent had Grade 3 arte-rial hypertension. The average proteinuria/creatininuria ratio was 3592.08 ± 7009.57 mg/g and 32% of women had glomerular grade proteinuria. The mean serum creatinine was 13.61 ± 12.62 mg/l. AKI (Acute Renal Failure) was present in 30% of women. All patients had received a central antihypertensive drug of which 88% were a calcium channel blocker. For the delivery mode, a Caesarean section was performed in 88% of cases. In the three months postpartum, 40% of women had persistent hypertension, 58% had persistent proteinuria and 6% had persistent impaired renal function. Prematurity (p = 0.0091), IUGR (intrauterine growth restriction) (p = 0.0012) and IUFD (intrauterine fetal death) (p = 0.0012) were associated with the persistence of proteinuria at M3 postpartum. Conclusion: Symptoms of pre-eclampsia do not automatically disappear after the delivery. Proteinuria and renal failure can persist beyond three months after the delivery and require treatment by a nephrologist.

Application of Pomegranate Rind Extract for Improvement of Functional Properties of Various Materials—A Critical Review

Pomegranate rind is abundantly available as a waste material. Pomegranate Rind Extract (PRE) can be applied to cotton fabrics for its natural colours, as a mordanting agent and also for imparting certain functional properties such as fire retardancy and antimicrobial properties. This paper reviews the feasibility of Pomegranate Rind Extract to improve the functional properties of cellulosic fabrics. Studies show that varying concentrations and higher temperatures that were used to apply the extract on the fabric, resulted in enhanced functional properties. At a particular concentration, the treated fabric showed a 15 times lower burning rate in comparison with the control fabric. Also, antimicrobial efficacy has been observed against Gram-positive and Gram-negative bacteria. Due to the natural colouring material, it can be used as a natural dye on cotton material. The fire retardancy of pomegranate rind extract was tested on jute material under varying alkalinity. Research has indicated that pomegranate rind extract could be used to dye polyamide as well. The rubbing and wash fastness of the finished fabrics is good. The light fastness was fair, and its antibacterial efficiency against tested bacteria was good.

Risk factors and prediction score for chronic pancreatitis: A nationwide population-based cohort study

AIM To explore the risk factors of developing chronic pancreatitis(CP) in patients with acute pancreatitis(AP) and develop a prediction score for CP.METHODS Using the National Health Insurance Research Database in Taiwan, we obtained large, population-based data of 5971 eligible patients diagnosed with AP from 2000 to 2013. After excluding patients with obstructive pancreatitis and biliary pancreatitis and those with a follow-up period of less than 1 year, we conducted a multivariate analysis using the data of 3739 patients to identify the risk factors of CP and subsequently develop a scoring system that could predict the development of CP in patients with AP. In addition, we validated the scoring system using a validation cohort.RESULTS Among the study subjects, 142 patients(12.98%) devel-oped CP among patients with RAP. On the other hand, only 32 patients(1.21%) developed CP among patients with only one episode of AP. The multivariate analysis revealed that the presence of recurrent AP(RAP), alcoholism, smoking habit, and age of onset of < 55 years were the four important risk factors for CP. We developed a scoring system(risk score 1 and risk score 2) from the derivation cohort by classifying the patients into low-risk, moderate-risk, and high-risk categories based on similar magnitudes of hazard and validated the performance using another validation cohort. Using the prediction score model, the area under the curve(AUC) [95% confidence interval(CI)] in predicting the 5-year CP incidence in risk score 1(without the number of AP episodes) was 0.83(0.79, 0.87), whereas the AUC(95%CI) in risk score 2(including the number of AP episodes) was 0.84(0.80, 0.88). This result demonstrated that the risk score 2 has somewhat better prediction performance than risk score 1. However, both of them had similar performance between the derivation and validation cohorts.CONCLUSION In the study,we identified the risk factors of CP and developed a prediction score model for CP.

Automatic Fault Prediction of Wind Turbine Main Bearing Based on SCADA Data and Artificial Neural Network

As the demand for wind energy continues to grow at exponential rate, reducing operation and maintenance (O & M) costs and improving reliability have become top priorities in wind turbine maintenance strategies. Prediction of wind turbine failures before they reach a catastrophic stage is critical to reduce the O & M cost due to unnecessary scheduled maintenance. A SCADA-data based condition monitoring system, which takes advantage of data already collected at the wind turbine controller, is a cost-effective way to monitor wind turbines for early warning of failures. This article proposes a methodology of fault prediction and automatically generating warning and alarm for wind turbine main bearings based on stored SCADA data using Artificial Neural Network (ANN). The ANN model of turbine main bearing normal behavior is established and then the deviation between estimated and actual values of the parameter is calculated. Furthermore, a method has been developed to generate early warning and alarm and avoid false warnings and alarms based on the deviation. In this way, wind farm operators are able to have enough time to plan maintenance, and thus, unanticipated downtime can be avoided and O & M costs can be reduced.

Relationship between variation of HBV preS function domains and clinical consequences of HBV infection

Objective:To investigate the relationship between variation of HBV preS genes and clinical consequencesof HBV infection. Methods: We selected 3 groups (3 in each) of HBV infected individuals including those of acuteinfection with complete recovery, chronic HBV carriers and chronic severe hepatitis B. The preS genes were amplified from serum samples with half nested PCR. The PCR products were cloned into M13 vectors and 10 cloneswere randomly selected for each individual and sequenced with standard methods. A total of 90 clones were sequenced and analyzed by DNA homological comparison. Results: The characteristics of the preS genes were significantly different in HBV-infected individuals with different clinical consequences.①As for acute hepatitis B, thepreS function domains were stable and no changes in epitopes for T and B cells were found. ②For chronic HBVcarriers, the function domains within preS proteins showed potentially changes with epitope drifts of T and B cellsbecause of the high variation of HBV preS genes. ③For chronic severe hepatitis B, there were special changes inthe function domains of B cells. The abnormally high immune response would cause severe liver injury in these individuals. Conclusion: This study provides experimental evidence for researches about the possible association ofvariation of preS1/S2 functional regions with clinical consequences.

Research on Wind Power Prediction Modeling Based on Adaptive Feature Entropy Fuzzy Clustering

Wind farm power prediction is proposed based on adaptive feature weight entropy fuzzy clustering algorithm.According to the fuzzy clustering method,a large number of historical data of a wind farm in Inner Mongolia are analyzed and classified.Model of adaptive entropy weight for clustering is built.Wind power prediction model based on adaptive entropy fuzzy clustering feature weights is built.Simulation results show that the proposed method could distinguish the abnormal data and forecast more accurately and compute fastly.

Effect of Pre-Gelatinization Conditions on the Total Oxalate Content and Techno-Functional Properties of Taro (Colocasia esculenta) Flour

Like most roots and tubers, taro (Colocasia esculenta) corms have a short shelf-life due to the high moisture content, which aggravates their post-harvest losses. They also contain high amounts of calcium oxalates, limiting their use in food applications. To help add value and diversify the use of taro corms as well as curb food losses, various strategies have been proposed, such processing of the corms into flour. This study aimed at evaluating the total oxalate content and techno-functional properties of taro flour as affected by the pre-gelatinization conditions (i.e., method and time). Pre-gelatinized taro flour was prepared by subjecting peeled and cleaned taro corms to roasting (190°C), boiling (100°C), and steaming (100°C) for 10 min, 20 min and 30 min, respectively, for each method, followed by drying at 55°C and milling. Generally, all the properties of flour were significantly affected by the pre-gelatinization conditions (P 0.05). The total oxalate content of the pre-gelatinized taro flour ranged from 33.26 to 76.90 mg/100g. Pre-gelatinization by boiling significantly reduced the oxalate content (56.7%), while roasting resulted in the least reduction (36.2%). The flour colour i.e. L*, hue, and chroma ranged from 38.47° - 70.30°, 42.64° - 69.43°, and 7.78° - 10.58°, respectively. Roasting resulted in flour with the largest L* (70.30°) and hue angle (69.43°). Boiling also resulted in flour with the highest bulk density (BD) (0.86 g/cm3) and the lowest water solubility index (WSI) (9.39%). Steamed flour had the highest water absorption index (WAI) (3.81 g/g), water holding capacity (WHC) (4.59 g/g), and swelling capacity (SC) (4.86 g/g). This study shows that pre-gelatinization (i.e. by boiling, steaming or roasting) significantly affects the total oxalate content and techno-functional properties of taro flour, which in turn influences its use in other food applications thus increasing the utilization and production of taro simultaneously.

Localizing the Language Network with fMRI and Functional Connectivity: Implications for Pre-Surgical Planning

Object: Functional MRI is frequently applied to lateralize language in pre-surgical planning, with potential to localize functionally important cortex too. Here we present BOLD signal activation maps and related functional connectivity, in response to three commonly administered fMRI language tasks. Methods: Datasets from 55 pre-surgical fMRI studies were analyzed. Verbal response naming, covert word generation and passive listening tasks were administered in all studies. Single-subject analyses, group analyses and region-of-interest analyses were conducted, and a multi-subject functional connectivity analysis was performed. Results: Single-subject analyses revealed that clinically important language regions were activated in all but three patients using the panel of tasks. Group analyses revealed significant bilateral BOLD signal increases in anterior and posterior language regions in response to verbal response naming and bilateral signal increase in posterior language regions only in response to passive listening. Covert word generation activated anterior language regions bilaterally and posterior language cortex in the dominant hemisphere. Functional connectivity analyses confirmed that activated regions were significantly correlated in all tasks. Conclusion: The findings of single-subject and group analyses add to the evidence supporting the use of a panel of fMRI tasks to map the language network for pre-surgical planning. Our findings support the additional use of functional connectivity analysis in routine language mapping to add to the localization value to fMRI. In addition, the results of our investigation demonstrate these three commonly applied tasks reliably activate unique aspects of the language network, which advocates closer individual inspection, guided by the surgical intervention planned.

pre-S基因变异与乙型肝炎病毒感染患者肝功能和肝纤维化的相关性

目的 研究了pre-S基因变异与乙型肝炎病毒(HBV)感染患者肝功能和肝纤维化的相关性。方法 选取2019年12月至2022年6月该院收治的HBV感染者178例为研究对象,其中无症状HBV携带患者65例作为对照组,慢性乙型肝炎(CHB)患者70例作为CHB组,乙型肝炎肝硬化患者43例作为肝硬化组。比较对照组、CHB组、肝硬化组患者的pre-S基因变异情况、肝功能和肝纤维化指标水平。根据是否发生基因缺失变异,将HBV感染患者分为变异组和非变异组,比较变异组和非变异组的肝功能和肝纤维化指标水平。采用Spearman相关分析pre-S基因变异类型与患者肝功能和肝纤维化指标的相关性。结果 对照组和CHB组pre-S基因缺失变异率明显低于肝硬化组,CHB组pre-S基因缺失变异率明显低于肝硬化组,差异均有统计学意义(P<0.05)。对照组和CHB组ALT、AST、层粘连蛋白、透明质酸以及Ⅲ型前胶原肽水平明显低于肝硬化组,清蛋白水平明显高于肝硬化组,差异均有统计学意义(P<0.05)。非变异组ALT、AST、层粘连蛋白、透明质酸以及Ⅲ型前胶原肽水平明显低于变异组,清蛋白水平明显高于变异组,差异均有统计学意义(P<0.05)。pre-S基因缺失变异与ALT、AST、层粘连蛋白、透明质酸、Ⅲ型前胶原肽水平均呈正相关(P<0.05),与清蛋白水平呈负相关(P<0.05)。结论 pre-S基因缺失变异与HBV感染患者肝功能和肝纤维化指标明显相关。

预康复在促进老年肺癌手术患者术后康复中的应用研究

目的研究预康复对老年肺癌手术患者肺功能、机体应激以及情感状况的影响。方法选择2020年10月至2022年12月于成都医学院第一附属医院行手术治疗的老年肺癌患者98例作为研究对象,依照随机数字表法分为预康复组(n=49)与对照组(n=49)。对照组行常规干预,预康复组行常规干预+预康复。观察两组胸管留置时间、术后1、3 d疼痛情况、总住院时间等手术相关指标。观察两组入院时、术前及术后1 h用力肺活量(FVC)、第1秒用力呼气容积(FEV1)及每分钟最大通气量(MVV)等肺功能指标;入院时、术前及术后7 d焦虑、抑郁评分;术后并发症。结果预康复组胸管留置时间、术后1、3 d疼痛情况、总住院时间均低于对照组(P<0.05)。术后1个月,预康复组FVC、FEV1、MVV均大于对照组(P<0.05)。术后7 d,预康复组焦虑、抑郁评分低于对照组(P<0.05)。预康复组术后并发症发生率低于对照组(P<0.05)。结论预康复可有效促进老年肺癌手术患者术后康复,提高其肺功能,改善其焦虑、抑郁情绪,对于改善患者预后具有重要作用。

糖尿病前期及2型糖尿病皮层萎缩与认知功能的相关性研究

目的探讨糖尿病前期(pre-diabetes mellitus,PDM)和2型糖尿病(type 2 diabetes mellitus,T2DM)大脑皮层的改变模式及其与认知功能的关系,以期探寻血糖代谢异常被试早期脑损害的影像标志物。材料与方法本研究纳入96例T2DM患者、30例PDM被试以及48名正常对照(normal control,NC),对被试进行了认知功能测试、临床生化检查及高分辨率3D-T1WI磁共振扫描。使用CAT12软件进行基于体素的形态学分析和基于表面的形态学分析得到全脑灰质体积和皮层厚度、局部回指数等皮层结构参数,并比较3组间的差异,结果均使用P<0.05的阈值和FWE校正进行多重比较校正。进一步提取组间具有差异的参数,与生化指标及认知量表得分进行相关分析。结果与NC相比,PDM被试右侧额下回眶部及左侧中央后回灰质萎缩(P<0.05,FWE校正),T2DM患者出现更多灰质萎缩,特别是右侧颞上回、右侧额下回眶部、右侧颞中回及左侧中央后回,右侧前额叶皮层厚度减小(P<0.05,FWE校正)。在血糖代谢异常被试中全脑灰质体积与胰岛素抵抗指数(r=−0.227,P=0.012,未校正)及连线测试A得分(r=−0.250,P=0.001,FDR校正)呈负相关,与数字广度-倒背得分呈正相关(r=0.267,P=0.003,FDR校正);皮层厚度与糖化血红蛋白(r=−0.181,P=0.040,未校正)及餐后2 h血糖(r=−0.272,P=0.020,未校正)呈负相关,与餐后2 h胰岛素(r=0.236,P=0.010,未校正)及胰岛β细胞功能指数(r=0.207,P=0.022,未校正)呈正相关。结论本研究发现PDM人群已存在脑区灰质萎缩,T2DM患者出现更多的灰质萎缩,且与注意和工作记忆功能相关,因此皮层萎缩有可能是糖尿病相关脑损伤早期的影像标志物。

血凝酶用于多发性创伤患者院前急救的价值分析

目的探讨血凝酶用于多发性创伤患者院前急救的价值。方法选取2019年4月至2022年4月北京急救中心收治的104例多发性创伤患者作为研究对象,根据患者是否应用血凝酶分为观察组(常规院前急救并应用血凝酶治疗,n=52)与对照组(常规院前急救治疗,n=52)。比较2组患者的止血时间、平均出血量、单位面积出血,比较2组患者治疗前后的血红蛋白(Hb)、红细胞压积(HCT)、血小板聚集率(PAR)、凝血酶原时间(PT)、部分凝血活酶时间(APTT)、纤维蛋白原(FIB)、D-二聚体(D-D)、抗凝血酶Ⅲ(AT-Ⅲ)、组织型纤溶酶原激活物(t-PA)、纤溶酶原激活物抑制剂-1(PAI-1),比较2组患者的并发症、不良反应。结果观察组止血时间、平均出血量、单位面积出血均少于对照组(P<0.05);治疗后2组Hb、HCT、PAR水平低于治疗前,观察组高于对照组(P<0.05);2组患者治疗前后,PT、APTT、FIB、D-D、PAI-1、AT-Ⅲ、t-PA水平比较差异无统计学意义(P>0.05)。治疗后,2组患者并发症总发生率比较差异无统计学意义(P>0.05),观察组病死率低于对照组(P<0.05);观察组不良反应发生率为5.77%,对照组不良反应发生率为0,差异无统计学意义(χ^(2)=3.089,P=0.079)。结论多发性创伤患者院前急救应用血凝酶可减少出血量,对患者的凝血和纤维蛋白溶解功能无影响,安全性良好,有利于患者预后的改善。

术前床上排尿训练联合集束化护理对心房颤动射频消融术患者卧位排尿功能的影响

目的研究心房颤动(以下称“房颤”)患者在射频消融术治疗期间接受术前床上排尿训练联合集束化护理干预的临床效果。方法选取2021年7月至2022年8月南昌大学第一附属医院收治的64例房颤患者作为研究对象,采用随机数字表法将其分成对照组(32例)和观察组(32例)。对照组开展常规射频护理,观察组开展术前床上排尿训练联合集束化护理干预。比较两组的手术操作时间、术后首次排尿时间、导尿次数、术后12 h残余尿量、护理前后心理状态评分、治疗依从性、护理满意度、术后膀胱功能分级、尿潴留发生情况。结果观察组的手术操作时间、术后首次排尿时间短于对照组,差异有统计学意义(P<0.05)。观察组的导尿次数和术后12 h残余尿量少于对照组,差异有统计学意义(P<0.05)。观察组的住院时间短于对照组,差异有统计学意义(P<0.05)。护理前,两组患者的心理状态评分比较,差异无统计学意义(P>0.05);护理后,观察组的心理状态评分低于对照组,差异有统计学意义(P<0.05)。观察组的治疗依从性高于对照组,差异有统计学意义(P<0.05)。观察组的护理总满意度高于对照组,差异有统计学意义(P<0.05)。术后,观察组的膀胱功能分级优于对照组,差异有统计学意义(P<0.05)。观察组的尿潴留发生率低于对照组,差异有统计学意义(P<0.05)。结论房颤患者在射频消融术治疗期间接受术前床上排尿训练联合集束化护理干预,能够帮助改善膀胱功能,保持良好心态,缩短治疗操作时间和恢复时间,减少导尿次数,防止尿潴留的发生,使治疗依从性和满意度得到显著提升。

功能锻炼衣在学龄前期上肢骨折患儿功能锻炼中的应用

目的:探讨功能锻炼衣在学龄前期上肢骨折患儿功能锻炼中的应用效果。方法:选取2022年1月—6月浙江省某三级甲等儿童医院骨科创伤中心收治的60例学龄前期上肢骨折患儿作为研究对象,按照住院号末位奇偶数随机分为试验组和对照组各30例。试验组在术后第1天穿带卡通图案可拆卸计数球体,显示屏可计数、带播报器的印有卡通人物形象的衣服(功能锻炼衣)进行功能锻炼,每天3次,每次100下,为期3个月。对照组穿普通衣物在相同时间进行传统握力球进行功能锻炼。比较两组术后2个月肘关节、尺桡关节的关节活动度、术后2周生活自理能力得分、上肢骨折相关并发症情况及家长满意度情况。结果:试验组患儿关节活动度、日常生活自理能力得分、家长满意度均高于对照组;上肢骨折后术后的肿胀程度低于对照组,差异均有统计学意义(P<0.05)。结论:带卡通计数球,球体可计数可拆卸的功能锻炼衣能提高学龄前期上肢骨折患儿关节活动度、日常生活自理能力、家长满意度,减轻学龄前期上肢骨折患儿术后上肢的肿胀程度。

基于5G专网的院前急救运作模式、体系对创伤失血性休克患者凝血功能及炎性因子的影响研究

目的研究基于5G专网的院前急救运作模式、体系对创伤失血性休克患者凝血功能及炎性因子的影响。方法应用医学研究随机数字表法,选取我院2020年8月—2023年8月收治的创伤失血性休克患者100例,按照治疗路径不同分为对照组和观察组,对照组给予常规救治模式治疗,观察组给予基于5G专网的院前急救运作模式、体系治疗,比较两组凝血功能、炎性因子以及抢救总有效率。结果观察组PT、APTT、TT、FIB指标水平均低于对照组,PLT水平高于对照组(均P<0.05);观察组TNF-α、IL-1、IL-6、CRP等炎性因子指标水平以及抢救总有效率均低于对照组(均P<0.05)。结论基于5G专网的院前急救运作模式、体系对创伤失血性休克能明显改善患者的凝血功能,降低炎性因子指标水平,升高抢救有效率。

功能冲突论视域下先秦冲突的军事考古学观察

科塞功能冲突论可为以古代军事遗存及行为为主要研究对象的军事考古学研究,提供不同的视角和思路。在简要梳理其主要观点的基础上,从先秦冲突的原因、影响和调控诸方面切入,探讨其对军事考古学研究的启示,以期深化古史研究及认识当今面临的问题。先秦冲突爆发的原因可主要分为物质性和非物质性两大类,其中物质性原因主要与各方对资源占有和(或)使用的规则存在异议有关,不同的异议可能会影响冲突的模式和新规则的制定,神权因素在早期社会和国家形态的冲突中也占有重要位置。先秦时期冲突和角逐对华夏文化圈、早期文化共同体以及早期国家的形成和初步发展在某种意义上有着较显著的激发作用,但其形式上的暴力性、残酷性和复杂性也在日益增长,春秋战国以降尤其如此。冲突双方的会盟誓诅、和亲、质物交换等敌意替代、转移或释放机制可一定程度地调控冲突,但往往因未彻底解决核心矛盾难以取得稳定长效,特别在神权和王权逐渐淡出盟誓核心后。而一些对外冲突的原因中,不排除存在因统治者不愿或不敢直面自身内部矛盾,试图将敌对情绪转移至外部的可能,属于激进极端的敌意替代举措,未及矛盾根本。在军事考古学研究中,应始终坚持和平导向,以史为鉴,为促进社会和谐稳定发展,推动人类命运共同体构建提供学科支持。

不同覆膜处理下藠头根际土壤真菌群落的响应特征

【目的】探讨不同覆膜措施对藠头根际土壤真菌群落结构、生态功能和物种间相互作用的影响,以期为生产上地膜的使用与选择提供理论依据和数据支持。【方法】在藠头收获期采集覆盖普通透明地膜、蓝色地膜以及不覆膜处理下的藠头根际土壤,利用高通量测序技术,并结合FUNGuild数据库、分子生态网络等分析方法对各处理下根际土壤真菌群落进行研究。【结果】覆膜处理降低了真菌群落多样性,但在蓝色地膜处理下富集了相对丰度更高的腐质霉属(Humicola)、篮状菌属(Talaromyces)、曲霉属(As⁃pergillus)等有益真菌菌群;从真菌分子生态网络可以看出,地膜覆盖虽使网络变得简化,但真菌群落对外部环境的抵抗能力有所提高,其中蓝色地膜较普通地膜表现更为稳定;此外,对各处理的真菌群落进行功能预测后发现,腐生型功能菌群的相对丰度在蓝色地膜下占比更高,促进了有机物的分解和养分循环。【结论】藠头根际土壤的真菌群落结构、生态功能和物种间相互关系在不同覆膜处理下表现出显著差异,其中,蓝色地膜处理下的真菌群落更有利于藠头的生长发育,但同时也应考虑覆膜措施引起病原型功能菌群的增加可能带来的潜在风险。

VA-ECMO患者院内死亡风险预测模型的构建及验证:一项多中心、回顾性、病例对照研究

目的 分析静脉-动脉体外膜肺氧合技术(VA-ECMO)患者死亡危险因素,构建并验证VA-ECMO患者院内死亡风险预测模型。方法 采用便利抽样,选取2015年1月~2022年1月广东省3家三甲综合医院ICU的302例VA-ECMO患者作为研究对象,随机分为建模组201例,验证组101例。运用单因素及多因素Logistic回归分析VA-ECMO患者死亡危险因素,构建VAECMO患者死亡风险预测模型并以列线图形式呈现。使用受试者工作特征曲线(ROC曲线)、校准曲线和临床决策曲线评价模型区分度、一致性及临床有效性。结果 预测VA-ECMO患者院内死亡风险的最终模型包括了高血压(OR=3.694,95%CI 1.582-8.621)、连续性肾脏替代治疗(OR=9.661,95%CI 4.103-22.745)、钠离子(OR=1.048,95%CI 1.003-1.095)、血红蛋白(OR=0.987,95%CI 0.977-0.998)。建模组预测模型的受试者工作特征曲线下面积AUC=0.829(95%CI 0.770-0.889),高于4个单独危险因素(AUC<0.800)、APACHE II评分[AUC=0.777(95%CI 0.714-0.840)]、SOFA评分的[AUC=0.721(95%CI 0.647-0.796)]。内部验证结果显示模型的受试者工作特征曲线下面积AUC为0.774(95%CI 0.679-0.869),拟合优度检验结果χ2=4.629,P>0.05。结论构建的VA-ECMO患者院内死亡风险预测模型区分度、校准度及临床有效性均较好,优于常用疾病严重程度评分系统,对评估重症患者疾病的严重程度及预后的风险水平有重要意义。

使用CNN(卷积神经网络)-LSTM(长短期记忆)联合神经网络预测盾构隧道施工引起的地面沉降

[目的]地铁盾构隧道施工会引起周围地面沉降,影响周围环境。传统地面沉降预测方法难以综合考虑沉降影响因素,对此,为提高地面沉降的预测精度,使用CNN(卷积神经网络)-LSTM(长短期记忆)联合神经网络,对盾构隧道施工引起的地面沉降进行预测。[方法]以某地铁施工区间地面沉降监测数据为研究对象,使用CNN对影响参数(压缩模量、黏聚力、内摩擦角、泊松比、土层厚度、隧道埋深和施工参数)与地面沉降监测值进行连接,使用LSTM神经网络对地面沉降进行分析,建立了基于CNN-LSTM联合神经网络的地面沉降预测模型,探讨了同时考虑多个因素对地面沉降预测值的影响。[结果及结论]使用CNN对地面沉降相关的影响参数特征提取效果较好;所建CNN-LSTM模型的准确率比单独使用LSTM模型的准确率提高了3%、比传统BP(反向传播)神经网络模型准确率提高了9%;所建CNN-LSTM模型,对单测点短时间地面沉降预测准确率达到93%,预测值与监测值吻合较好。