Search

Background Recently, several studies have reported an association between lipid profiles and sudden sensorineural hearing loss (SSNHL), yet there is considerable variability between the individual studies in defining the precise association between serum lipids levels and SSNHL. This study sought to identify a possible relationship between dyslipidemia and the prevalence and prognosis of SSNHL. Methods A case‐control study was carried out at two independent medical centers, including 2,288 SSNHL patients and 2,288 healthy controls. Clinical characteristics and serum lipid parameters were assessed, including total cholesterol (CHOL), high‐density lipoprotein (HDL), low‐density lipoprotein (LDL), triglycerides (Trig), apolipoprotein AI (ApoAI), apolipoprotein B (ApoB), and lipoprotein a (Lpa). Multivariate logistic regression analysis was performed to assess the relationship between lipid profiles and SSNHL in the 4,576 subjects. Results Significant differences were identified in several conventional serum lipid markers including CHOL, Trig, HDL, LDL, ApoAI, ApoB, and Lpa, between SSNHL patients and healthy controls. Serum ApoAI levels were significantly lower in patients with bilateral SSNHL compared to unilateral SSNHL. Binary logistic regression analysis revealed that higher levels of ApoB, LDL, Trig, and lower levels of ApoAI and HDL were all associated with an increased risk of SSNHL. After clinical characterization, multivariate analysis showed that only low levels of ApoB predicted likelihood of a recovery of more than 30 dB among patients with SSNHL. Conclusions Serum lipids are associated with the incidence and prognosis of SSNHL. Identification of dyslipidemia may improve early evaluation and management of SSNHL risks., Binary logistic regression analysis of various lipids and SSNHL risk. Although CHOL and Lpa concentrations differed significantly between SSNHL patients and control subjects, neither the CHOL nor Lpa was related to SSNHL in multivariate analysis. Subjects with ApoB levels 5.54‐fold (95% CI 4.11–7.47) higher than that of healthy controls were more likely to have SSNHL pathology. Patients with abnormal LDL and Trig levels had a 10% (1.10 (1.03–1.18)) and 29% (1.29 (1.22–1.35)) increased odds of having SSNHL compared with healthy controls. Levels of ApoAI in patients with SSNHL were significantly lower (0.16 (0.13–0.21)) than that of control subjects. Conversely, HDL levels of control subjects were lower than those observed in patients with SSNHL (0.67 (0.56–0.80)).

Clustering has been widely used in the fields of knowledge discovery, pattern recognition and artificial intelligence. However, discovering clusters in spatial databases is still a challenging task, especially when the shape, size, and density of clusters vary a lot. Existing algorithms have sensitive parameters, clusters must be separated far enough from each other and rich prior knowledge about datasets is required. In this paper, we propose algorithm DENSS, which performs clustering on the basis of the similarity of neighbour distribution and the number of shared neighbors for two objects. Algorithm DENSS can mine clusters that differ in densities, and within a cluster the local densities are reasonably homogeneous. Adjacent objects are separated into different clusters by significant change in densities. To verify the effectiveness of the algorithm DENSS, synthetic and real-world datasets are used for testing, and it has been compared with seven clustering algorithms. Experimental results show that the proposed algorithm has a relatively high efficiency, robustness and effectiveness, and is remarkably superior to the seven algorithms. This algorithm is universal and can rapidly and efficiently identify the clusters of different densities, shapes and sizes even in the presence of noise and outliers for any object feature types.

In recent years, the technology about IoT (Internet of Things) has been applied into finance domain, and the generated data, such as the real-time data of chattel mortgage supervision with GPS, sensors, network cameras, mobile devices, etc., has been used to improve the capability of financial credit risk management of bank loans. Financial credit risk is by far one of the most significant risks that commercial banks have to face, however, when confronting to the massively growing financial data from multiple sources including Internet, mobile networks or IoT, traditional statistical models and neural network models might not operate fairly or accurately enough for credit risk assessment with those diverse data. Hence, there is a practical need to establish more powerful risk prediction models with artificial intelligence based on big data analytics to predict default behaviors with better accuracy and capacity. In this article, a big data mining approach of Particle Swarm Optimization (PSO) based Backpropagation (BP) neural network is proposed for financial risk management in commercial banks with IoT deployment, which constructs a nonlinear parallel optimization model with Apache Spark and Hadoop HDFS techniques on the dataset of on-balance sheet item and off-balance sheet item. The experiment results indicate that this parallel risk management model has fast convergence rate and powerful predictive capacity, and performs efficiently in screening default behaviors. In the meanwhile, the distributed implementation on big data clusters largely reduces the processing time of model training and testing.

OBJECTIVES To explore the risk factors on contrast agent venous intravasation during transvaginal 4-dimensional hysterosalpingo-contrast sonography (TVS 4D-HyCoSy). METHODS The TVS 4D-HyCoSy imaging data were collected from 276 female infertile patients. The correlation between endometrial thickness, days after menstruation, intrauterine intervention history, fallopian tubal patency degree, and contrast agent venous intravasation, respectively, was analyzed. RESULTS In our study, the incidence of contrast agent venous intravasation was 13.04%. Endometrial thickness and days after menstruation were significantly associated with venous intravasation (P

In the era of big data, rich information about the movement behaviors of residents can be easily obtained. The movement patterns of residents can be mined from these data, which is essential for researching the correlation among zones in cities, thereby shedding light upon the understanding of the travel needs of residents. Many clustering algorithms are proposed to mine meaningful movement patterns between zones from the massive amount of origin–destination flows. However, most existing flow clustering algorithms focus only on the spatial information of the flow, but ignore the time attributes, precluding the identification of when the movement pattern occurs and how long it lasts. Nevertheless, for most existing algorithms that separately consider time attributes and cluster time and space, large amounts of redundant calculations result in high time complexity. Furthermore, only a few studies have considered flow clustering; most of people-related flows in the city are restricted by the road network, thereby causing the analyzed movement pattern between zones to not be consistent with reality. To address these problems, the present paper proposes a novel spatiotemporal similarity measure algorithm for flows. The algorithm uses flows as basic units, considers the spatiotemporal attributes, and introduces the road network constraints to make the results further realistic, in a bid to mine network-constrained movement patterns featured by spatiotemporal characteristics. To verify the feasibility and effectiveness of the proposed algorithm, we conduct experiments using artificial and real-world datasets of commuting data in Shanghai. Experiment results show that the algorithm can quickly and effectively mine movement patterns between zones with spatiotemporal attributes.

In order to detect multiple sclerosis (MS) subjects from healthy controls (HCs) in magnetic resonance imaging, we developed a new system based on machine learning. The MS imaging data was downloaded from the eHealth laboratory at the University of Cyprus, and the HC imaging data was scanned in our local hospital with volunteers enrolled from community advertisement. Inter-scan normalization was employed to remove the gray-level difference. We adjust the misclassification costs to alleviate the effect of unbalanced class distribution to the classification performance. We utilized two-level stationary wavelet entropy (SWE) to extract features from brain images. Then, we compared three machine learning based classifiers: the decision tree, k-nearest neighbors (kNN), and support vector machine. The experimental results showed the kNN performed the best among all three classifiers. In addition, the proposed SWE+kNN approach is superior to four state-of-the-art approaches. Our proposed MS detection approach is effective.

Finding an appropriate and accurate technology for early detection of disease is significantly important to research early treatments. We proposed some novel automatic classification systems based on the stationary wavelet transform (SWT) and the improved support vector machine (SVM). Magnetic Resonance Imaging (MRI) is commonly used for brain imaging as a non-invasive diagnostic tool to assist the pre-clinical diagnosis. However, MRI generates a large information set, which poses a challenge for classification. To deal with this problem we proposed a new approach, which combines SWT and Principal Component Analysis for feature extraction. In our experiments, three different datasets and four kinds of classifiers of the SVM were employed. The results over 5×6-fold stratified cross-validation (SCV) for Dataset-66, and 5×5-fold SCV for the other two datasets show that the average accuracy is almost 100.00%.

An accurate diagnosis is important for the medical treatment of patients suffering from brain diseases. Magnetic resonance (MR) images are commonly used by technicians to assist preclinical diagnosis. The classification of MR images of normal and pathological brains poses a challenge from the technological point of view, since MR imaging generates a large information set that reflects the conditions of the brain. In this paper, we present a computer-assisted diagnosis method based on wavelet entropy (WE) of the feature space approach and a feed-forward neural network (FNN) classification method for improving the brain diagnosis accuracy by means of MR images. The most relevant image feature is selected as the WE, which is used to train an FNN classifier. The results using tenfold cross validation of 64 images show that the average accuracy attainable is 100.00%. It can be easily seen from the data that the proposed classifier can detect abnormal brains from normal controls with excellent performance, which can compete with the latest methods. © 2016 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

Emotion recognition represents the position and motion of facial muscles. It contributes significantly in many fields. Current approaches have not obtained good results. This paper aimed to propose a new emotion recognition system based on facial expression images. We enrolled 20 subjects and let each subject pose seven different emotions: happy, sadness, surprise, anger, disgust, fear, and neutral. Afterward, we employed biorthogonal wavelet entropy to extract multiscale features, and used fuzzy multiclass support vector machine to be the classifier. The stratified cross validation was employed as a strict validation model. The statistical analysis showed our method achieved an overall accuracy of 96.77±0.10%. Besides, our method is superior to three state-of-the-art methods. In all, this proposed method is efficient.

The theoretical analysis of drag torques within a two-speed dual clutch transmission is presented in this article. The numerical models are developed to study the different sources of drag torques in dual clutch transmission. Simulations are performed in Matlab/Simulink platform to investigate the variation of drag torques under different operating conditions. Then an experimental investigation is conducted to evaluate the proposed model using an electric vehicle powertrain test rig. Outcomes of experimentation confirm that simulation results agree well with test data. Therefore the proposed model performs well in the prediction of drag torque for the transmission, and can be applied to assess the efficiency of the transmission. Results demonstrate that the entire drag torque is dominated by the viscous shear in the wet clutch pack and gear churning losses. This lays a theoretical foundation to future research on reducing drag torque and applications of drag torque in powertrain system efficiency optimization.

When the single earth-return line of HVDC (High voltage direct current) system is functioning, the AC system in the same area will probably be influenced. Because two converting station at different DC potential constitute a loop via transmission line, causing certain direct current moving into the transformer neutral point and winding. The transformer DC magnetic biasing phenomenon will appear, leading increased noise and vibration of transformer. The influence of transformer DC magnetic biasing already exists and will do harm to transformers in a long term, while there's no set of effective quality inspection method. Therefore, it is necessary to establish a set of real-time current monitoring system for transformer DC magnetic biasing. In this paper, a set of automatic, high-accuracy and real-time current monitoring system is proposed. The monitoring system can display the amplitude and polarity of main transformer's neutral point direct current.

This paper first briefly introduces the basic structure of converter transformer, analyses the mechanism of vibration through simulation of the converter transformer oil-tank, designs the test scheme. This paper introduces the application of wavelet analysis method in converter transformer fault identification, then apply these methods to 24 converter transformers to obtain the characteristics parameters, finally utilize the vibro-acoustic techniques to diagnose power transformers. Through the experimental study of converter transformers in ±800kV Fengxian HVDC Substation combining with the existing criteria to evaluate existing converter transformer, and put forward the quantitative evaluation criteria.

This work aims at developing a novel pathological brain detection system (PBDS) to assist neuroradiologists to interpret magnetic resonance (MR) brain images. We simplify this problem as recognizing pathological brains from healthy brains. First, 12 fractional Fourier entropy (FRFE) features were extracted from each brain image. Next, we submit those features to a multi-layer perceptron (MLP) classifier. Two improvements were proposed for MLP. One improvement is the pruning technique that determines the optimal hidden neuron number. We compared three pruning techniques: dynamic pruning (DP), Bayesian detection boundaries (BDB), and Kappa coefficient (KC). The other improvement is to use the adaptive real-coded biogeography-based optimization (ARCBBO) to train the biases and weights of MLP. The experiments showed that the proposed FRFE?+?KC-MLP?+?ARCBBO achieved an average accuracy of 99.53 % based on 10 repetitions of K-fold cross validation, which was better than 11 recent PBDS methods.

Interaction among different regions can be illustrated in the form of a stream. For example, the interaction between the flows of people and information among different regions can reflect city network structures, as well as city functions and interconnections. The popularization of big data has facilitated the acquisition of flow data for various types of individuals. The application of the regional interaction model, which is based on the summary level of individual flow data mining, is currently a hot research topic. Thus far, however, previous research on spatial interaction methods has mainly focused on point-to-point and area-to-area interaction patterns, and investigations on the patterns of interaction hotspots between two regional groups with predefined neighborhood relationships, that being with two regions, remain scarce. In this study, a method for the identification of similar interaction hotspot patterns between two regional groups is proposed, and geo-information Tupu methods are applied to visualize interaction patterns. China’s air traffic flow data are used as an example to illustrate the performance of the proposed method to identify and analyze interaction hotspot patterns between regional groups with adjoining relationships across China. Research results indicate that the proposed method efficiently identifies the patterns of interaction flow hotspots between regional groups. Moreover, it can be applied to analyze any flow space in the excavation of the patterns of regional group interaction hotspots.

In order to study the economy performance of a two-speed electric vehicle (EV) equipped with a wet dual clutch transmission (DCT), this paper presents a theoretical analysis of drag torque in wet clutches, and investigates the relationship between drag torque, transmission gear ratios, and their optimisation. As such, the application of drag torque effects in gear ratio optimisation is conducted. Through this paper mathematic models for a two-speed EV is presented and a procedure to obtain the reasonable optimal gear ratio pairs and corresponding shift schedule for a two-speed EV is described in detail, with consideration of wet clutch drag torque. Through simulation and parametric analysis using different test driving cycles, including constant speed, NEDC and UDDS, the effects of drag torque is studied in gear ratio selection and EV efficiency and economy, with result demonstrating a significant influence on overall vehicle performance, and degradation of drivetrain efficiency.

(Aim) Tea plays a significant role because of its high value throughout the world. Computer vision techniques were successfully employed for rapid identification of teas. (Method) In our work, we present a computer assisted discrimination system on the basis of two steps: (i) two-dimensional wavelet-entropy for feature extraction; (ii) the feedforward Neural Network (FNN) for classification. Specifically, the wavelet entropy features were fed into a FNN classifier. (Results) The 10 runs of 75 images of three categories showed that the average accuracy achieved 90.70 %. The sensitivities of green, Oolong, and black tea are 92.80 %, 84.60 %, and 96.30 %, respectively. (Conclusions) It was easily observed that the proposed classifier can distinguish tea categories with satisfying performances, which was competitive with recent existing systems.

It is important to early detect pathological brains. Traditional methods used plain support vector machine (SVM) that is vulnerable to noises and outliers. In this study, we presented a hybrid method that combined wavelet-energy (WE) and fuzzy support vector machine (FSVM). The results over a 5x5-fold cross validation showed that the proposed "WE + FSVM" produced accuracy of 93.78%, higher than "WE + KSVM" of 91.78%, "DWT + PCA + RBF-NN" of 91.33%, "WE + BP-NN" of 86.67%, and "DWT + PCA + BP-NN" of 86.22%. Therefore, this study offered a new means to solve the problem with excellent performance.

An accurate diagnosis is important for the medical treatment of patients suffered from brain disease. Nuclear magnetic resonance images are commonly used by technicians to assist the pre-clinical diagnosis, rating them by visual evaluations. The classification of NMR images of normal and pathological brains poses a challenge from technological point of view, since NMR imaging generates a large information set that reflects the conditions of the brain. In this work, we present a computer assisted diagnosis method based on a wavelet-entropy (In this paper 2D-discrete wavelet transform has been used, in that it can extract more information) of the feature space approach and a Naive Bayes classifier classification method for improving the brain diagnosis accuracy by means of NMR images. The most relevant image feature is selected as the wavelet entropy, which is used to train a Naive Bayes classifier. The results over 64 images show that the sensitivity of the classifier is as high as 94.50%, the specificity 91.70%, the overall accuracy 92.60%. It is easily observed from the data that the proposed classifier can detect abnormal brains from normal controls within excellent performance, which is competitive with latest existing methods.

As pure electric vehicles are considered to be a major growth trend in the automotive industry, research into and development of efficient electric powertrain systems and related control technologies have become popular research topics. The growing importance and use of multi-speed transmissions in these vehicles make shift schedule design and research a crucial aspect of the powertrain systems design of pure electric vehicles. This paper provides a gear shift schedule calculation method for pure electric vehicles, which includes a dynamic shift schedule and an economic shift schedule calculation method, demonstrating how to optimize the shift points and to produce the upshift and downshift lines based on the motor efficiency map. Through the establishment of a pure electric vehicle model, simulation results show that a properly designed shift schedule can improve the working region of the motor and can refine the dynamic performance and the economic performance of the vehicle. Finally, rig testing results are demonstrated to be comparable with simulations and indicate the correctness of the method.

This paper proposes an approach to optimize the economy performance of a two-speed electric vehicle (EV) by combining gear shifting schedule design and gear ratios selection. Mathematic models for the two-speed EV subsystems are developed, including those of the battery module, electric machine, the driver, transmission and vehicle. Then a procedure for obtaining the optimal gear ratio pairs and corresponding shift schedule for the two-speed EV is presented in detail. The optimized EV powertrain parameters can not only ensure that basic requirements in dynamic performance are achieved, but realize the optimal economic performance of the EV as well. In order to investigate the effectiveness of the proposed method for EV design, simulations based on the developed powertrain model is conducted using different test driving cycles, including NEDC and constant speed. Results of these simulations validate the effectiveness of the proposed optimization method. Copyright © 2013 SAE International.

Method to improve the efficiency and the design of more efficient pure electric vehicle (PEV) system is a research hotspot for the automobile industry worldwide. Traditionally, PEV powertrains are dominated almost exclusively by single reducer structures in both concept and commercially available PEVs. This paper presents a novel two motor two speed pure electric power-train structure to study alternate methods for improving the operating efficiency of a typical PEV. This idea is derived from conventional power splitting transmissions to achieve motor efficiency optimization and ultimately lead to running range improvement. Through parameter matching and simulation, results demonstrated that the proposed two motor two speed PEV can realize better vehicle performance than single reducer PEV system; furthermore it can behave even better than a similar single motor two speed system. Copyright © 2013 SAE International.

© The Author(s) 2017. This article develops a systematic model to study electric vehicle powertrain system efficiency by combining a detailed model of two-speed dual-clutch transmission system efficiency losses with an electric vehicle powertrain system model. In this model, the design factors including selection of the electric machine, gear ratios' change, multi-plate wet clutch design, and gear shift schedule design are considered. Meanwhile, the application of detailed model for drag torque losses in the gearbox is discussed. Furthermore, the proposed model, developed with the MATLAB/Simulink platform, is applied to optimize/maximize the efficiency of the electric vehicle powertrain system using genetic algorithms. The optimization results demonstrate that the optimal results are different between simulations via New Europe Drive Cycle and Urban Dynamometer Driving Schedule, and comprehensive design and optimization of the powertrain system are necessary.

Dual clutch transmissions (DCTs) are recognized as being suitable for electric drive applications as they can drive with high efficiency and achieve good shifting comfort. A two-speed DCT electric drivetrain is described in this paper, comprised of only two gear pairs and a final drive gear in the two-speed gearbox. The fundamental shifting control algorithm is provided. On the testing rig of University of Technology, Sydney (UTS) powertrain lab, shifting controls and some driving cycle controls were realized. The results demonstrated that the control algorithm functioned well both in transient shifting control process and in the driving cycle conditions. © 2013 Bo Zhu et al.