Clinical impact of endoscopy position detecting unit(UPD-3) for a non-sedated colonoscopy

AIM:To evaluate whether an endoscopy position detecting unit(UPD-3) can improve cecal intubation rates, cecal intubation times and visual analog scale(VAS) pain scores, regardless of the colonoscopist's level of experience.METHODS:A total of 260 patients(170 men and 90women)who underwent a colonoscopy were divided into the UPD-3-guided group or the conventional group(no UPD-3 guidance).Colonoscopies were performed by experts(experience of more than 1000colonoscopies)or trainees(experience of less than 100colonoscopies).Cecal intubation rates,cecal intubation times,insertion methods(straight insertion:shortening the colonic fold through the bending technique;roping insertion:right turn shortening technique)and patient discomfort were assessed.Patient discomfort during the endoscope insertion was scored by the VAS that was divided into 6 degrees of pain.RESULTS:The cecum intubation rates,cecal intubation times,number of cecal intubations that were performed in<15 min and insertion methods were not significantly different between the conventional group and the UPD-3-guided group.The number of patients who experienced pain during the insertion was markedly less in the UPD-3-guided group than in the conventional group.Univariate and multivariate analysis showed that the following factors were associated with lower VAS pain scores during endoscope insertion:insertion method(straight insertion)and UPD-3guidance in the trainee group.For the experts group,univariate analysis showed that only the insertion method(straight insertion)was associated with lower VAS pain scores.CONCLUSION:Although UPD-3 guidance did not shorten intubation times,it resulted in less patient painduring endoscope insertion compared with conventional endoscopy for the procedures performed by trainees.

AN ALGORITHM OF MOTION ESTIMATION BASED ON UNIT QUATERNION DECOMPOSITION OF THE ROTATION MATRIX

Based on the unit quaternion decomposition of rotation matrix, this paper puts forward an algorithm to estimate motion parameters from the space position vectors of 3D feature points. Rotation matrix’s representation with the unit quaternion has no singular points, so the unit quaternion-based estimation method is of more practical importance, and the algorithm in this paper does not need iteration computation compared to those unit quaternion-based methods proposed by Horn(1987) and Su, et al.(1989). Solution’s uniqueness analysis of the algorithm and simulation experiment results are also presented, it can be seen that performance of our method is satisfactory.

Role of proning and positive end-expiratory pressure in COVID-19

The novel coronavirus,which was declared a pandemic by the World Health Organization in early 2020 has brought with itself major morbidity and mortality.It has increased hospital occupancy,heralded economic turmoil,and the rapid transmission and community spread have added to the burden of the virus.Most of the patients are admitted to the intensive care unit(ICU)for acute hypoxic respiratory failure often secondary to acute respiratory distress syndrome(ARDS).Based on the limited data available,there have been different opinions about the respiratory mechanics of the ARDS caused by coronavirus disease 2019(COVID-19).Our article provides an insight into COVID-19 pathophysiology and how it differs from typical ARDS.Based on these differences,our article explains the different approach to ventilation in COVID-19 ARDS compared to typical ARDS.We critically analyze the role of positive end-expiratory pressure(PEEP)and proning in the ICU patients.Through the limited data and clinical experience are available,we believe that early proning in COVID-19 patients improves oxygenation and optimal PEEP should be titrated based on individual lung compliance.

Coping with the Impact of Climate Change: A Dive into Precision Agriculture in the United States

With the continued increase in the number of people that are food insecure globally, which could be increasing because of the ongoing Ukraine-Russia war, leading to reduction in international agribusinesses, coupled with drastic climate change exacerbating the problem of food insecurity, there is a constant need to come up with innovative approaches to solve this global issue. In this article, we articulated how precision agriculture can be a tool for ensuring food security in the United States. This study aims to reiterate the significance of precision agriculture in solving global food insecurity.

改进PU-GAN的点云上采样网络

针对点云分类问题,提出一种改进点云上采样网络,从低密度点云生成密集点云,提高点云分类准确率。针对点云上采样网络PU-GAN细节表示能力不足的问题,在生成器网络中引入Transformer模块,整合提取的特征信息;在特征扩张模块加入门控循环单元和分层上采样单元,重构细粒度特征;以PU-GAN数据集进行训练,构建闽南古建筑数据集作为测试。实验结果表明,改进后网络的上采样效果获得了提升,具有良好的鲁棒性。通过对ModelNet40数据集进行上采样,在PointNet上进行分类实验,验证了该网络对分类准确率的提升。

一种基于聚类的PU主动文本分类方法

文本分类是信息检索的关键问题之一.提取更多的可信反例和构造准确高效的分类器是PU(positive and unlabeled)文本分类的两个重要问题.然而,在现有的可信反例提取方法中,很多方法提取的可信反例数量较少,构建的分类器质量有待提高.分别针对这两个重要步骤提供了一种基于聚类的半监督主动分类方法.与传统的反例提取方法不同,利用聚类技术和正例文档应与反例文档共享尽可能少的特征项这一特点,从未标识数据集中尽可能多地移除正例,从而可以获得更多的可信反例.结合SVM主动学习和改进的Rocchio构建分类器,并采用改进的TFIDF(term frequency inverse document frequency)进行特征提取,可以显著提高分类的准确度.分别在3个不同的数据集中测试了分类结果(RCV1,Reuters-21578,20 Newsgoups).实验结果表明,基于聚类寻找可信反例可以在保持较低错误率的情况下获取更多的可信反例,而且主动学习方法的引入也显著提升了分类精度.

3-PU^＊U^＊型平动并联机构的运动学分析

研究 3-PU U 型平动并联机构的运动学分析问题。通过将该机构简化成运动等效的机构模型 ,得到其位置逆解公式及描述机构输入与输出速度关系的逆雅可比 ,并进一步对 3-PU U 并联机构作业空间的几何性质、机构的运动灵活性进行了分析 ,给出了尺度参数变化对机构作业空间及运动灵活性的影响规律。分析表明 ,该机构具有几何形状规则的作业空间及较好的运动灵活性 。

基于PU学习算法的虚假评论识别研究

识别虚假评论有着重要的理论意义与现实价值.先前工作集中于启发式策略和传统的全监督学习算法.最近研究表明:人类无法通过先验知识有效识别虚假评论,手工标注的数据集必定存在一定数量的误例,因此简单使用传统的全监督学习算法识别虚假评论并不合理.容易被错误标注的样例称为间谍样例,如何确定这些样例的类别标签将直接影响分类器的性能.基于少量的真实评论和大量的未标注评论,提出一种创新的PU(positive and unlabeled)学习框架来识别虚假评论.首先,从无标注数据集中识别出少量可信度较高的负例.其次,通过整合LDA(latent Dirichlet allocation)和K-means,分别计算出多个代表性的正例和负例.接着,基于狄利克雷过程混合模型(Dirichlet process mixture model,DPMM),对所有间谍样例进行聚类,混合种群性和个体性策略来确定间谍样例的类别标签.最后,多核学习算法被用来训练最终的分类器.数值实验证实了所提算法的有效性,超过当前的基准.

基于支持向量机的PU中文文本分类器构建

在PU文本分类中,针对文本分类中可信反例获取困难,使文本分类器性能低下的问题,提出了一种基于支持向量机的PU文本分类器的构建。该方法采用改进的特征词权值提取方法进行特征词提取,利用OB_PCZ算法从未标识文本数据集中尽可能多地移除正例,从而获得更多的可信反例。结合支持向量机和改进的Rocchio构建了高效的文本分类器。实验表明,该方法比目前其它的面向PU学习的文本分类方法具有更高的分类质量。

Purex流程铀钚分离工艺单元数学模型研究

在铀钚分离工艺单元单级数学模型和混合澄清槽瞬态数学模型的基础上,建立了以U(Ⅳ)-N_2H_4为还原反萃剂、混合澄清槽为萃取设备的Purex流程铀钚分离工艺单元数学模型,开发了计算机模拟程序,并使用台架实验数据对程序的可靠性进行了验证。结果表明,模拟程序的计算值和实验值符合良好。在此基础上,利用模拟软件对铀钚分离工艺单元的工艺参数进行了计算分析,结果表明:1BX_1加入位置、1BS和1BX_2酸度对钚反萃率无太大影响,但1BX_1加入位置和补萃级数对钚中去铀系数SF_(U/Pu)有一定影响。

3-UPU转动并联机构的两类位置正解分析

根据纯转动型 3 UPU并联机构的几何特点 ,使用四元数变换方法 ,构造了一类单正交条件下的 3 UPU并联机构位置正解分析的参数化方程 ,可使用三角恒等变换来求解该方程 ;并使用单位四元数变换求出其双正交条件下的 8个封闭的位置正解 ,该正解为

硫磺回收装置急冷塔衬PU防腐技术的应用

介绍了炼油厂硫磺回收装置急冷塔采用内衬PU材料进行防腐的应用实例。对比以往急冷塔壳体使用特材或碳钢内衬PO等材料,明显缩短了急冷塔制造工期,也减少了在材料上的投资成本。急冷塔采用碳钢内衬PU后,后续检修工作方便快捷,缩短了开车停工时间。列出了PU材料的主要性能和优点,介绍了急冷塔壳体用碳钢衬PU的制造施工方法和合格检验方法。

基于PU学习的建议语句分类方法

建议挖掘作为一项新兴研究任务,具有重要的应用价值。针对传统建议语句分类方法所存在的规则复杂、标注工作量大、特征维度高、数据稀疏等问题,提出一种基于PU学习的建议语句分类方法。首先,使用简单规则从无标注评论集合中选择建议语句的正例集合;然后,为了降低特征维度,缓解数据稀疏性,在自编码神经网络(Autoencoder)特征空间中使用Spy技术划分可靠反例集合;最后,利用正例集合和可靠反例集合来训练多层感知机(MLP)对剩余的无标注样例进行分类。该方法在中文数据集上的F1值和准确率值分别达到81.98%和82.67%,实验结果表明,该方法能够有效地对建议语句进行分类,且不需要对数据进行人工标注。

基于Copula函数的列车定位惯性单元剩余寿命预测方法

列车连续、高精度定位需要惯性测量单元(IMU)提供可靠的姿态校正补偿数据,IMU输出的数据质量会受外部环境干扰、内部结构磨损等影响而降低,因此有必要对IMU性能指标参数及剩余使用寿命(RUL)预测开展研究。根据IMU组成结构与性能评价准则,选取IMU中陀螺仪和加速度计的随机误差指标表征设备性能状态,在对IMU多关键部件特征进行分析的基础上,设计基于非线性漂移Wiener过程的IMU性能退化模型,构建基于Copula函数的二元性能参量RUL联合分布函数;分别采用后向平滑(RTS)和期望最大化(EM)联合算法、极大似然估计(MLE)算法估计性能退化模型参数和Copula函数中未知参数,实现RUL的实时自适应更新预测。基于RUL预测方法对IMU剩余使用寿命进行预测,与列车定位系统中IMU的实际性能退化监测数据对比。结果表明,采用本文提出的算法,IMU失效前最后6个监测时刻RUL预测期望值的均方相对误差(MRE)为9.87%,相比基于陀螺仪寿命预测方法降低了3.27%。

DTU-PU:针对不确定数据PU学习的决策树

不确定数据的PU学习在现实世界的许多应用中,如在传感器网络、市场分析和医学诊断等领域普遍存在,提出了针对不确定数据PU学习的决策树算法。基于POSC45中信息增益的计算方法,引入UDT中处理连续属性的不确定数据时用到的不确定数据区间及概率分布函数的概念,提出了一种能处理连续属性的不确定数据PU学习的决策树算法DTU-PU(Decision Tree for Uncertain data with PU-learning)。在UCI数据集上的实验表明,DTU-PU具有较好的分类准确率和健壮性。

An indoor positioning system for mobile target tracking based on VLC and IMU fusion

An indoor positioning system( IPS) is designed to realize positioning and tracking of mobile targets,by taking advantages of both the visible light communication( VLC) and inertial measurement unit( IMU). The platform of the IPS is designed,which consists of the light-emitting diode( LED)based transmitter,the receiver and the positioning server. To reduce the impact caused by measurement errors,both inertial sensing data and the received signal strength( RSS) from the VLC are calibrated. Then,a practical propagation model is established to obtain the distance between the transmitter and the receiver from the RSS measurements. Furthermore,a hybrid positioning algorithm is proposed by using the adaptive Kalman filter( AKF) and the weighted least squares( WLS)trilateration to estimate the positions of the mobile targets.Experimental results show that the developed IPS using the proposed hybrid positioning algorithm can extend the localization area of VLC,mitigate the IMU drifts and improve the positioning accuracy of mobile targets.

结合半监督与主动学习的时间序列PU问题分类

目前基于PU问题的时间序列分类常采用半监督学习对未标注数据集U中数据进行自动标注并构建分类器,但在这种方法中,边界数据样本类别的自动标注难以保证正确性,从而导致构建分类器的效果不佳。针对以上问题,提出一种采用主动学习对未标注数据集U中数据进行人工标注从而构建分类器的方法 OAL(Only Active Learning),基于投票委员会(QBC)对标注数据集构建多个分类器进行投票,以计算未标注数据样本的类别不一致性,并综合考虑数据样本的分布密度,计算数据样本的信息量,作为主动学习的数据选择策略。鉴于人工标注数据量有限,在上述OAL方法的基础上,将主动学习与半监督学习相结合,即在主动学习迭代过程中,将类别一致性高的部分数据样本自动标注,以增加训练数据中标注数据量,保证构建分类器的训练数据量。实验表明了该方法通过部分人工标注,相比半监督学习,能够为PU数据集构建更高准确率的分类器。

基于四元数理论的2UPS-PU并联机构位置反解

介绍了四元数在描述刚体位置和方位中的应用,并简要描述了2UPS-PU并联机构,接着给出了利用四元数计算2UPS-PU并联机构的位置反解的推导过程,得到了2UPS-PU并联机构的位置反解公式,最后结合MATLAB和SolidWorks Motion进行仿真分析,仿真结果验证了该位置反解方法的正确性和有效性。

Finite Element Analysis of Vehicle Load Effect on Harvesting Energy Properties of a Piezoelectric Unit

To realize the goal of harvesting energy from pavement vibration on a large scale, a new type of piezoelectric harvesting units as the energy transducer has been proposed. The piezoelectric harvesting units are paved 40 mm below the asphalt, which is the same as thickness of the top layer of typical asphalt pavement in China. The spacing distance is 2200 mm, which is the same as the one between two tires of a normal vehicle. A mathematical model of the unit is deduced on Meda empirical formula and Hamilton principle and piezoelectric equations. Effects of the external vehicle load on its harvesting energy properties and pavement deformation and stress are analyzed with the finite element method. The results show that the excited voltage is linearly variation with contact pressure while the harvested electrical energy exponential varies with contact pressure. The more the contact pressure is, the larger the harvested electrical energy and the deformation and stress are. The harvested electrical energy also increases with the load frequency. At least 100 mJ of electrical energy can be collected with the proposed piezoelectric harvesting unit. It shows that the technology application of the piezoelectric harvesting energy from pavement is promising.

Effects of the prone position on arterial blood gas analysis and respiratory parameters of acute respiratory distress syndrome patients:An observational retrospective study

Objective:To investigate the effect of the first prone position on arterial blood gas analysis and respiratory parameters of acute respiratory distress syndrome(ARDS)patients with and without COVID.Methods:This study was conducted retrospectively with 22 COVID-ARDS and 22 non-COVID ARDS patients,who were placed in a prone position for at least 16 hours on the first day at the intensive care unit admission,and arterial blood gas analysis was taken in the pre-prone,prone and post-prone periods.Results:PaO2 were significantly increased in the pre-prone vs.prone comparison in both groups,but the increase in the PaO2/FiO2 ratio was not significant.In comparing the pre-prone vs.post-prone PaO2/FiO2 ratios,there was a significant difference only in the non-COVID ARDS group.Conclusions:The improved oxygenation provided by prone positioning is more permanent with the“post-prone effect”in non-COVID ARDS patients.This can be attributed to the differences in the pathogenesis of the two ARDS types.