Search

The microscale combustion calorimeter (MCC) is considered to be the most commonly used small-scale fire experiment in material research. It was developed by Richard E. Lyon and Richard Walters from the Federal Aviation Administration and has existed for over two decades. During this period, the apparatus has been used to obtain combustion properties for a wide range of materials as well as screening flame retardants for the production of fire-resistant materials. MCC separately simulates the pyrolysis and combustion stages of a burning process in a non-flaming test and uniquely measures the heat release capacity. Since its development, several significant modifications have been made in the apparatus to rectify some of the sources of error. This research reviews the literature on microscale combustion calorimetry. It covers the principles of operation applications and limitations, prediction methods and mathematical models for estimation as well as the modifications made in the design. Furthermore, the relationship between MCC parameters and measurement from other fire experiments are outlined. Lastly, the experimental methods designed on the basis of microscale combustion calorimetry have been briefly described.

L Antigen Family Member 3 (LAGE3) is an important RNA modification-related protein. Whereas few studies have interrogated the LAGE3 protein, there is limited data on its role in tumors. Here, we analyzed and profiled the LAGE3 protein in skin cutaneous melanoma (CM) using TCGA, GTEx, or GEO databases. Our data showed an upregulation of LAGE3 in melanoma cell lines compared to normal skin cell lines. Besides, the Kaplan–Meier curves and Cox proportional hazard model revealed that LAGE3 was an independent survival indicator for CM, especially in metastatic CM. Moreover, LAGE3 was negatively associated with multiple immune cell infiltration levels in CM, especially CD8+ T cells in metastatic CM. Taken together, our study suggests that LAGE3 could be a potential prognostic biomarker and might be a potential target for the development of novel CM treatment strategies.

Most of the current works are still limited to dealing with the melody generation containing pitch, rhythm, duration of each note, and pause between notes. This paper proposes a transformer-based model to generate multi-track music including tracks of piano, guitar, and drum, which is abbreviated as MTMG model. The proposed MTMG model is mainly innovated and improved on the basis of transformer. Firstly, the model obtains three target sequences after pairwise learning through learning network. Then, according to these three target sequences, GPT is applied to predict and generate three closely related sequences of instrument tracks. Finally, the three generated instrument tracks are fused to obtain multi-track music pieces containing piano, guitar, and drum. To verify the effectiveness of the proposed model, related experiments are conducted on a pair of comparative subjective and objective evaluation. The encouraging performance of the proposed model over other state-of-the-art models demonstrates its superiority in musical representation.

Automatic music composition could dramatically decrease music production costs, lower the threshold for the non-professionals to compose as well as improve the efficiency of music creation. In this paper, we proposed an intelligent music composition neutral network to automatically generate a specific style of music. The advantage of our model is the innovative structure: we obtained the music sequence through an actor’s long short term memory, then fixed the probability of sequence by a reward-based procedure which serves as feedback to improve the performance of music composition. The music theoretical rule is introduced to constrain the style of generated music. We also utilized a subjective validation in experiment to guarantee the superiority of our model compared with state-of-the-art works.

Image denoising is an important technology for image preprocessing. In recent years, the image denoising technology based on total variation (TV) has been rapidly developed. However, However, although it can preserve image details well, which generates obvious staircase effects. This is due to the traditional TV-based image denoising technology only applies the gradient information and ignored the local variance of the image. In order to suppress staircase effect, in this paper, a novel image denoising approach based on TV model and weighting function is proposed. First, the theory mechanism of staircase effect brought by the traditional TV model is analyzed. Second, the effects of weighting function on edge regions, flat regions, and gradation and detail regions are also analyzed. Third, based on the above analysis, an improved TV model is proposed. Finally, the image denoising approach is implemented by an iterative algorithm. The experimental results show that, compared with various state-of-the-art models denoising models, the proposed image denoising approach can effectively suppress the staircase effect of the traditional TV model in most cases, preserve the image details, and improve the image denoising performance.

On May 18, 2006, a disastrous groundwater inrush occurred at the Xinjing coal mine in Datong, Shanxi province, China. Great effort was taken during the post-accident rescue. However, triggered by the accumulated water rushed from the nearby abandoned tunnels and goafs of the Yanzishan coal mine, which had been closed for tens of years, it caused great damage, including 56 deaths and direct economic loss of over 53 million yuan. The outburst groundwater was from the abandoned goafs in the No.14-1 coal seam of the neighbour Yanzishan coal mine. The passage formed in northeast part of the mining area in the No. 14-1 coal seam. The average inflow rate was 23,333.3 m3/h. Unidentified spatial distribution and water-filled degree of the abandoned tunnels and goafs of the Yanzishan coal mine are critical contributions to the accident. Illegal mining in the No.14-1 coal seam is an anthropogenic contribution. That mandatory regulations for excessive groundwater exploration and release were not carried out in the mining is the third fatal cause leading to the accident. Finally, the poor awareness on water inrush recognition and control of the miners also induced the accident to a disastrous extent.

Existing outdoor three-dimensional (3D) object detection algorithms mainly use a single type of sensor, for example, only using a monocular camera or radar point cloud. However, camera sensors are affected by light and lose depth information. When scanning a distant object or an occluded object, the data collected by the short-range radar point cloud sensor are very sparse, which affects the detection algorithm. To address the above challenges, we design a deep learning network that can combine the texture information of two-dimensional (2D) data and the geometric information of 3D data for object detection. To solve the problem of a single sensor, we use a reverse mapping layer and an aggregation layer to combine the texture information of RGB data with the geometric information of point cloud data and design a maximum pooling layer to deal with the input of multi-view cameras. In addition, to solve the defects of the 3D object detection algorithm based on the region proposal network (RPN) method, we use the Hough voting algorithm implemented by a deep neural network to suggest objects. Experimental results show that our algorithm has a 1.06% decrease in average precision (AP) compared to PointRCNN in easy car object detection, but our algorithm requires 37.7% less time to calculate than PointRCNN under the same hardware environment. Moreover, our algorithm improves the AP by 1.14% compared to PointRCNN in hard car object detection.

Deep blue fluorescent materials are crucial in the commercialization of organic light-emitting diodes (OLEDs) for full-color displays or solid-state lighting sources. In this work, a series of aromatic ring compounds, based on a newly designed 2-(pyridine-4-yl)-3-phenyl-1H-benzo[g]indole core, have been synthesized, on which donor (D) and acceptor (A) groups are bonded in a Ψ-type configuration, forming a D–π–A–π–D structure. The electronic structure and photophysical properties have been explored, as well as the applications in the blue OLEDs. The results show that multistate couplings exist between the three D–π–A–π–D moieties, resulting in a high photoluminescence quantum yield up to 76.01% peaking at around 410 nm, and a depressed efficiency roll-off at high luminance. The deep blue OLEDs based on CzCNBPyIp emitter exhibits stable emission peaking at 416 nm with a negligible efficiency roll-off at a high luminance range of 1 000–10 000 cd m−2, and corresponding CIEx,y = (0.162, 0.085) and maximum EQE = 2.6%, making it one of the highest performing solution-processed deep blue fluorescent devices. This work provides an alternative method for the deep blue fluorescent material design toward solution processing OLEDs.

Predictions of both combustible material flammability and heat release parameters have been long goals in fire safety research, for its complex heat, mass transfer and chemical reaction process in gas phase. In this study, neural network method is employed to predict materials flammability considering its wide application in predicting key properties of engineering problems. The use of group method of data handling (GMDH) and feed forward back-propagation (FFBP) neural networks in predicting the heat of combustion and heat release capacity (HRC) from microscale combustion calorimetry has been examined. The study presented models with excellent predictability though GMDH out-performed FFBP. The deviation of the predicted and measured HRC data from this study was compared with the results of other predictive modelling techniques used in flammability studies. The GMDH neural network results presented the least mean deviation of 4.01 signifying accurate predictions. Hence, this study proposed the use of GMDH in predicting flammability characteristics of materials.

Currently, multi-label automatic image annotation (MAIA) approach based on machine learning has been widely applied and developed. Since extreme learning machine (ELM) has the advantages of simple structure, fast learning speed, better generalisation ability and so on, it is used for MAIA in this study. In order to enhance the annotation performance and generalisation ability of MAIA, some work is designed and implemented. First of all, a novel distance metric learning method based on cost-sensitive learning for MAIA task is proposed to reduce the impact of class imbalance of samples. Second, an improved ELM approach based on singular value decomposition is proposed for implementing MAIA task. Finally, the selection of training samples (STS) strategy based on error correlation is also proposed to improve the generalisation ability and annotation performance of MAIA. Based on the above work, a novel MAIA approach is implemented. The experimental results confirm that the proposed cost-sensitive DLM, improved cost-sensitive ELM and STS can obtain the good generalisation ability, and achieve better annotation performance than the existing MAIA approaches.

Automated image annotation (AIA) is an important issue in computer vision and pattern recognition, and plays an extremely important role in retrieving large-scale images. In many image annotation approaches, different regions of the image are processed equally, which is inconsistent with the mechanism by which humans understand images. In order to improve the annotation performance of existing AIA approaches, a hybrid AIA approach based on visual attention mechanism (VAM) and the conditional random field (CRF) is proposed. First, since people pay more attention to the salient region of an image during the image recognition process, VAM is implemented for acquiring the salient and non salient regions of the image. Second, support vector machine (SVM) is used to annotate the salient region, and k nearest neighbor (kNN) voting algorithm is used to annotate the non salient regions. Finally, due to the existence of a certain relationship between any two annotation words (also called labels), CRF is calculated to obtain the final label set of each given image. The experimental results confirm that the proposed hybrid AIA approach has ideal annotation performance.

Recently, deep neural networks have exhibited as a powerful architecture to well capture the nonlinear distribution of high-dimensional multimedia data such as image, video, text and audio, so naturally does for multi-modal data. How to make full use of multimedia data? This leads to an important research direction: cross-modal learning. In this paper, we introduce a method based on the content of audio and video data modalities implemented with a novel two-branch neural network is to learn the joint embeddings from a shared subspace for computing the similarity between the two modalities. In particular, the contribution of proposed method is mainly manifested in the three aspects: i) Using feature selection model for choosing top-k audio and visual feature representation; ii) A novel combination of training loss function concerning inter-modal similarity and intra-modal invariance is used; iii) Due to the lack of video-music paired dataset, we construct dataset of video-music pairs from YouTube 8M and MER31K datasets. The experiments have proved that our proposed model has a better performance compared with other methods.

The impact of buried air gaps (voids) in SiO 2 cladding in common silicon photonics building blocks is investigated through simulation. The void cross-sectional areas in devices fabricated through a commercial electron beam lithography process are evaluated and found to be less than 500 nm2.

Background Recent infection testing algorithms (RITAs) incorporating clinical information with the HIV recency assay have been proven to accurately classify recent infection. However, little evidence exists on whether RITAs would help in precisely identifying major drivers of the ongoing HIV epidemic. Methods HIV recency test results and clinical information were collected from 1152 newly diagnosed HIV cases between 2015 and 2017 in Dehong prefecture of Yunnan province, and the efficacy of four different RITAs in identifying risk factors for new HIV infection was compared. Results RITA 1 uses the recency test only. RITA 2 and RITA 3 combine the recency test with CD4+ T cell count and viral load (VL), respectively. RITA 4 combines both CD4+ T cell count and VL. All RITAs identified the MSM group and young people between 15 and 24 years as risk factors for incident HIV infection. RITA 3 and RITA 4 further identified the Dai ethnic minority as a risk factor, which had not been identified before when only the HIV recency test was used. Conclusions By comparing different RITAs, we determined that greater accuracy in classifying recent HIV infection could help elucidate major drivers impacting the ongoing epidemic and thus inform targeted interventions.

A modified total variation MRI image denoising method is proposed in this paper. First, the proposed method removes the noise in K-space in compressed sensing MRI reconstruction. Then, the removed K-space data is used as a partial frequency observation in compressed sensing MRI model. The proposed method shows better results than RecPF method, LDP method, TVCMRI method, and FCSA method in sparse MRI reconstruction. The proposed method is tested against Shepp-Logan phantom and real MR images corrupted by noise of different intensity level, and it gives better Signal-to-Noise Ratio (SNR), the relative error (ReErr), and the structural similarity (SSIM) than RecPF, LDP, TVCMRI, and FCSA., Medical imaging; Mathematics; Total variation denoising; K-space data; MRI reconstruction; Compressed sensing

With the continuous development of deep learning, research on the conversion of singing voice has gradually enriched. The study of singing voice conversion comes from voice conversion. The singing voice conversion is to sing the voice of the target singer without changing the sound content of the source singer. In this paper, we use WORLD vocoder and speech signal processing toolkit (SPTK) to extract the acoustic characteristics of songs and use two mirrored generative adversarial Nets complete the conversion of acoustic features. The experiment realizes the singing conversion of non-parallel corpus. Subjective evaluations show that the songs after the conversion have a good performance in the quality and similarity of the songs.

The segmentation of buildings in remote-sensing (RS) images plays an important role in monitoring landscape changes. Quantification of these changes can be used to balance economic and environmental benefits and most importantly, to support the sustainable urban development. Deep learning has been upgrading the techniques for RS image analysis. However, it requires a large-scale data set for hyper-parameter optimization. To address this issue, the concept of &ldquo, one view per city&rdquo, is proposed and it explores the use of one RS image for parameter settings with the purpose of handling the rest images of the same city by the trained model. The proposal of this concept comes from the observation that buildings of a same city in single-source RS images demonstrate similar intensity distributions. To verify the feasibility, a proof-of-concept study is conducted and five fully convolutional networks are evaluated on five cities in the Inria Aerial Image Labeling database. Experimental results suggest that the concept can be explored to decrease the number of images for model training and it enables us to achieve competitive performance in buildings segmentation with decreased time consumption. Based on model optimization and universal image representation, it is full of potential to improve the segmentation performance, to enhance the generalization capacity, and to extend the application of the concept in RS image analysis.

Deep learning techniques applied in image style transfer have shown great success. In the domain of music, this task can further aid musicians and music-generating AIs in the production of music. In this paper we present an unsupervised methodology for musical style transfer, which do not rely on strictly paired data to train. The model is capable of translating the style of symbolic music from the source domain to the target domain while mostly preserving the content and structure of input data. Output samples are evaluated by genre classifiers and show promising results.

How to retrieve the desired images quickly and accurately from the large scale image database has become a hot topic in the field of multimedia research. Many content-based image retrieval (CBIR) technologies already exist, but they are not always satisfactory. In many applications, the CBIR model based on machine learning relies heavily on the distance metric between samples. Although the traditional distance metric methods are simple and convenient, it is not always appropriate for CBIR tasks. In this paper, a novel distance metric learning (DML) method based on cost sensitive learning (CSL) is studied, and then it is used in a large margin distribution learning machine (LDM) to replace the traditional kernel functions. The improved LDM also takes into account CSL, and which is called CS-DLDM. Finally, CS-DLDM model is applied to CBIR tasks for implementation classification. We compare the proposed CS-DLDM model with other classifiers based on CSL. The experimental results show that the proposed CS-DLDM model not only has satisfactory classification performance but also the lowest misclassification cost, can effectively avoid the class imbalance of sample.

Many personalized advertisement recommendation studies suffer from the problem of only certain tagged items can be recommended in video playback, which mean it can’t recommend more produces to users that they really like . It also doesn’t know the users really like at the source. Due to the large number of scene changes in different video, the users can choose more items they like. This study attempts to adopt transfer knowledge to solve the problem of data volume to provide users with a variety of options. Aiming at the image classification model of learning on big data set, this paper proposes a method to solve the problem of scene object recognition in TV program,such as movies,TV plays, variety shows and short video, by transferring a pre-trained depth image classification model to a specific task. In a small training set, Learning high-level representations on a small training set to produce a task-specific target model. Experiments on small data sets and real face sets collected by myself show that the transfer learning is effective and efficient. In the application of video, this study provides a theoretical basis for personalized click recommendation of video users.

The small-world architecture has gained considerable attention in anatomical brain connectivity studies. However, how to adequately quantify small-worldness in diffusion networks has remained a problem. We addressed the limits of small-world measures and defined new metric indices: the small-world efficiency (SWE) and the small-world angle (SWA), both based on the tradeoff between high global and local efficiency. To confirm the validity of the new indices, we examined the behavior of SWE and SWA of networks based on the Watts-Strogatz model as well as the diffusion tensor imaging (DTI) data from 75 healthy old subjects (aged 50-70). We found that SWE could classify the subjects into different age groups, and was correlated with individual performance on the WAIS-IV test. Moreover, to evaluate the sensitivity of the proposed measures to network, two network attack strategies were applied. Our results indicate that the new indices outperform their predecessors in the analysis of DTI data.

Background: Schistosomiasis is an immunopathogenic disease in which Th17 cells play vital roles. Hepatic granuloma formation and subsequent fibrosis are its main pathologic manifestations and the leading causes of hepatic cirrhosis, and effective therapeutic interventions are lacking. In this study, we explored the effects of fasudil, a selective RhoA&ndash, Rho-associated kinase (ROCK) inhibitor, on Th17 cells and the pathogenesis of schistosomiasis. Methods: Mice were infected with Schistosoma japonicum and treated with fasudil. The worm burden, hepatic granuloma formation, and fibrosis were evaluated. The roles of fasudil on Th17, Treg, and hepatic stellate cells were analyzed. Results: Fasudil therapy markedly reduced the granuloma size and collagen deposit in livers from mice infected with S. japonicum. However, fasudil therapy did not affect the worm burden in infected mice. The underlying cellular and molecular mechanisms were investigated. Fasudil suppressed the activation and induced the apoptosis of CD4+ T cells. Fasudil inhibited the differentiation and effector cytokine secretion of Th17 cells, whereas it upregulated Treg cells in vitro. It also restrained the in vivo interleukin (IL)-4 and IL-17 levels in infected mice. Fasudil directly induced the apoptosis of hepatic stellate cells and downregulated the expressions of hepatic fibrogenic genes, such as collagen type I (Col-I), Col-III, and transforming growth factor-1 (TGF-&beta, 1). These effects may contribute to its anti-pathogenic roles in schistosomiasis. Conclusions: Fasudil inhibits hepatic granuloma formation and fibrosis with downregulation of Th17 cells. Fasudil might serve as a novel therapeutic agent for hepatic fibrosis due to schistosome infections and perhaps other disorders.

Background: Outflow tract (OFT) septation defects are a common cause of congenital heart disease. Numerous studies have focused on the septation mechanism of the OFT, but have reported inconsistent conclusions. This study, therefore, aimed to investigate the septation of the aortic sac and the OFT in the early embryonic human heart. Methods: Serial sections of 27 human embryonic hearts from Carnegie stage (CS) 10 to CS19 were immunohistochemically stained with antibodies against c-smooth muscle actin (α-SMA) and myosin heavy chain. Results: At CS10–CS11, the OFT wall was an exclusively myocardial structure that was continuous with the aortic sac at the margin of the pericardial cavity. From CS13 onward, the OFT was divided into nonmyocardial and myocardial portions. The cushion formed gradually, and its distal border with the OFT myocardium was consistently maintained. The aortic sac between the fourth and sixth aortic arch arteries was degenerated. At CS16, the α-SMA-positive aortopulmonary septum formed and fused with the two OFT cushions, thus septating the nonmyocardial portion of the OFT into two arteries. At this stage, the cushions were not fused. At CS19, the bilateral cushions were fused to septate the myocardial portion of the OFT. Conclusions: Data suggest that the OFT cushion is formed before the aortopulmonary septum is formed. Thus, the OFT cushion is not derived from the aortopulmonary septum. In addition, the nonmyocardial part of the OFT is septated into the aorta and pulmonary trunk by the aortopulmonary septum, while the main part of the cushion fuses and septates the myocardial portion of the OFT. Key words: Aortopulmonary Septum; Human Embryonic Heart; Immunohistochemistry; Outflow Tract; Outflow Tract Cushion

Background HIV-1 viral load (VL) testing is essential for monitoring the effects of antiretroviral therapy in HIV-infected patients. In order to identify factors that impact testing performance of HIV-1 VL testing laboratories, this study performed a retrospective analysis on data from the domestic HIV-1 VL proficiency testing (PT) program in China during 2005 to 2019. Methods Analysis was conducted on testing results of 155 PT panel specimens that were distributed to HIV-1 VL testing laboratories nationwide during 2005 to 2019. Follow-up on-site inspection records on unqualified laboratories were also analyzed. Results By 2019, 267 HIV-1 VL testing laboratories in China enrolled in the national PT assessment. Overall, HIV-1 VL testing performance has been consistently good after 2012, with less than 5% of participants reporting an unqualified score. Unsatisfactory equipment conditions and poor laboratory administration were the two main reasons causing unqualified scores in the PT assessment. Interestingly, we found that HIV-1 VL testing laboratories had better performance in the CDC system than in hospitals. In analysis on the variance of specimen testing results by different assays, we found that variation in results existed across different assay platforms, and HIV-1 VL testing assays based on real-time PCR technology showed comparatively smaller inter-laboratory variability. Conclusions To maintain high performance in HIV-1 VL testing laboratories, it is important to strengthen laboratory management and preserve equipment conditions. Due to the variation of testing results among different assay platforms, we suggest using the same assay platform for longitudinal monitoring of patient viral load.

Software defect prediction (SDP) is very helpful for optimizing the resource allocation of software testing and improving the quality of software products. The cross-project defect prediction (CPDP) model based on machine learning is first learned through the existing training data with sufficient number and defect labels on one project, and then used to predict the defect labels of another new project with insufficient number and fewer labeled data. However, its prediction performance has a large gap compared with the within-project defect prediction (WPDP) model. The main reason is that there are usually differences between the distributions of training data in different software projects, and it has a greater impact on the prediction performance of the CPDP model. To solve this problem, the kernel twin support vector machines (KTSVMs) is used to implement domain adaptation (DA) to match the distributions of training data for different projects. Moreover, KTSVMs with DA function (called DA-KTSVM) is further used as the CPDP model in this paper. Since the parameters of DA-KTSVM have an impact on its predictive performance, these parameters are optimized by an improved quantum particle swarm optimization algorithm (IQPSO), and the optimized DA-KTSVM is called as DA-KTSVMO. In order to confirm the effectiveness of DA-KTSVMO, some experiments are implemented on 17 open source software projects. Experimental results and analysis show that DA-KTSVMO can not only achieve better prediction performance than other CPDP models compared, but also achieve almost the same or better compared performance than WPDP models when the training sample data is sufficient. In addition, DA-KTSVMO can make better use of existing sufficient data knowledge and realize the reuse of defective data to improve the prediction performance of DA-KTSVMO.

Relatively little attention has been paid to ADP-ribosylated modifications of histones, especially to mono-ADP-ribosylation. As an increasing number of mono-ADP-ribosyltransferases have been identified in recent studies, the functions of mono-ADP-ribosylated proteins have aroused research interest. In particular, histones are substrates of some mono-ADP-ribosyltransferases and mono-ADP-ribosylated histone have been detected in physiological or pathological processes. In this research, arginine-117 (Arg-117; R-117) of hsitone3(H3) is identified as the a site of mono-ADP-ribosylation in colon carcinoma(the first such site to be identified); this posttranslational modification may promote the proliferation of colon carcinoma cells in vitro and in vivo. Using a point-mutant lentivirus transfection and using an activator of P300 allowed us to observe the mono-ADP-ribosylation at H3R117 and enhancement of the activity of P300 to up-regulate the level of acetylated β-catenin, which could increase the expression of c-myc and cyclin D1.

STUDY DESIGN Surgical approach development in an animal model, and a prospective study comparing clinical outcomes between novel and conventional approaches in thoracolumbar burst fracture fixation. OBJECTIVE To investigate the feasibility of a less-invasive retroperitoneal approach to the lumbar spine in a sheep model and to compare the clinical outcomes of anterior reconstruction in the treatment of thoracolumbar burst fractures using novel and conventional approaches. SUMMARY OF BACKGROUND DATA The anterior retroperitoneal lumbar approach is well established for anterior lumbar surgical procedures in both humans and animal models. However, potential concerns include the increased risk of complications such as soft-tissue trauma, and extended periods of rehabilitation postoperatively. MATERIALS AND METHODS A less-invasive retroperitoneal approach was designed in a sheep model with minimal soft-tissue dissection to keep the abdominal and paravertebral muscles intact. Eight sheep underwent anterior lumbar interbody fusion using this approach. In the clinical study, 48 patients with thoracolumbar burst fractures underwent anterior decompression and reconstruction. The less-invasive approach and conventional approach were applied in 12 and 36 cases, respectively. The clinical outcomes during the minimum 12-month follow-up of the 2 groups were compared. RESULTS With the less-invasive approach, anterior lumbar interbody fusion was accomplished in all sheep, and no surgical complications were observed. In the clinical study, operation time, blood loss, and duration of hospitalization were comparable between 2 groups. Using the less-invasive approach decreased the length of incision, 3-day postoperative visual analogue scale score, postoperative independent standing, and narcotic-dependent duration. No surgical complications were observed in either group. CONCLUSIONS Our results and early experience suggests that the less-invasive retroperitoneal approach is safe and effective for anterior lumbar surgery. Compared with the conventional approach, significantly better postoperative rehabilitation and abdominal muscle preservation were seen with this novel approach.

Hubs are a set of highly connected brain regions which play an important role in the human brain network. Various graph measurements and parcellation atlases have been utilized for hub identification. However, the relationship of those measurements and the comparison of hub identification results derived from different types of brain networks are not clear yet. The current study used four measurements to identify hubs in 75 healthy aging subjects’ brain networks. Also, to figure out the potential effect using various parcellation schemes on the hub identification, five kinds of brain networks were constructed, which were accomplished by two types of brain parcellation schemes including anatomical parcellation atlases (AAL atlas and HOA altas) and random parcellation scheme (uniformed parcellation atlas with 32, 128 and 512 regions). From the results, we found that hubs can be consistently identified in the same type of brain network regardless of measurements. On the contrary, hubs were notably different in the different types of brain networks even using the same measurement. Beyond these, hub consistency between measurements derived from anatomical brain networks tend to be relatively stable than that derived from uniformed parcellated brain networks. Importantly, the consistency of identification results of uniformed parcellated brain networks were also constrained by the selection of identification threshold level. For the relationship between graph measurements, results revealed a robust relationship between vulnerability and other measurements. Our findings provide a better understanding to the effect of identification measurements, parcellation atlases, and identification thresholds on the hub identification, which may offer a future prospect of being able to create a unified standard for the hub identification.

Background This study assessed clinical and radiographic outcomes of oblique lumbar interbody fusion (OLIF) in comparison with posterior reoperation for adjacent segment disease (ASD). Methods A total of 26 patients with symptomatic ASD after lumbar fusion were included in this retrospective case-controlled study conducted from January 2013 to December 2018. Twelve patients underwent single-segment OLIF with or without posterior instrumentation (OLIF group), whereas 14 patients underwent posterior reoperation (posterior approach group). The clinical outcomes included operative time, blood loss, hospital stay, Visual Analogue Scale (VAS), Oswestry Disability Index (ODI), and complications. Preoperative and postoperative radiographic outcomes were compared. Results The operative time (60.6 ± 16.1 min vs. 150.9 ± 28.5 min, respectively; P

The Internet of Things (IoT) is a burgeoning paradigm that changes our lives greatly. In many cases, IoT devices are located in remote areas, which can’t be served by terrestrial access networks. As a consequence, Low Earth Orbit (LEO) satellite communication system may play an important evolution of 5G, and becomes of paramount importance for those scenes relying on its disadvantages. Additionally, the first integration of terrestrial communication and satellite communication in 5G also provides more technical support for the 5G system. Based on the analysis of the special application scenarios and traffic distribution characteristics of LEO satellite based IoT, this paper describes a simulation method for the traffic of LEO satellite based IoT. The simulation result shows that the distribution of business in LEO satellite system has great suddenness and variability both in time and space. Since the non-uniformity is not conducive to the stability of the system, it is necessary to guide the construction of LEO satellite based IoTs according to the distribution characteristics and laws of the business.

Music classification is a challenging task in music information retrieval. In this article, we compare the performance of the two types of models. The first category is classified by Support Vector Machine (SVM). We use the feature extraction from audio as the basis of classification. Firstly, a total of 500 pieces of music by five famous classical music composers were selected, 400 of which were regarded as the training set of music genre classification, and the remaining pieces were regarded as the testing set. The second method is Multilevel Wasserstein Means(MWMs). From the experimental results, Multilevel Wasserstein Means, as an unsupervised learning, has been able to approach SVM in classification results，and has achieved 85% accuracy.

Background Why some persons living with human immunodeficiency virus (HIV) (PLWH) progress quickly and others remain “healthy” for a decade or more without treatment remains a fundamental question of HIV pathology. We aimed to assess the epidemiological characteristics of HIV long-term nonprogressors (LTNPs) based on a cohort of PLWH in China observed between 1989 and 2016. Methods We conducted a nationwide, retrospective cohort study among Chinese PLWH with HIV diagnosed before 1 January 2008. Records were extracted from China’s national HIV/AIDS database on 30 June 2016. LTNPs were defined as those with AIDS-free, antiretroviral therapy–naive survival, with CD4 cell counts consistently ≥500/μL for ≥8 years after diagnosis. Prevalence was calculated, characteristics were described, and determinants were assessed by means of logistic regression. Potential sources of bias were also investigated. Results Our cohort included 89 201 participants, of whom 1749 (2.0%) were categorized as LTNPs. The injection drug use (IDU) route of infection was reported by 70.7% of LTNPs, compared with only 37.1% of non-LTNPs. The odds of LTNP status were greater among those infected via IDU (adjusted odds ratio [95% confidence interval], 2.28 [1.94–2.68]) and with HIV diagnosed in settings with large populations of persons who inject drugs (1.75 [1.51–2.02] for detention centers, 1.61 [1.39–1.87] for Yunnan, 1.94 [1.62–2.31] for Guangdong, and 2.90 [2.09–4.02] for Xinjiang). Conclusions Overrepresentation of the IDU route of infection among LTNPs is a surprising finding worthy of further study, and this newly defined cohort may be particularly well suited to exploration of the molecular biological mechanisms underlying HIV long-term nonprogression., Persons who acquired human immunodeficiency virus (HIV) via injection drug use are surprisingly overrepresented in the largest nationwide Chinese cohort of human HIV long-term nonprogressors (LTNPs), suggesting that further investigation is warranted and may elucidate mechanisms underlying LTNP phenotypes.

Scientists have used many different classification methods to solve the problem of music classification. But the efficiency of each classification is different. In this paper, we propose two compared methods on the task of music style classification. More specifically, feature extraction for representing timbral texture, rhythmic content and pitch content are proposed. Comparative evaluations on performances of two classifiers were conducted for music classification with different styles. The result shows that XGB is better suited for small datasets than BPNN, Comment: 5 pages, 1 figures

Autopilot is the latest application of artificial intelligence technology in the automotive field. This is the future trend of automotive development and is seen as a beacon to lead the new automotive industry revolution. There are many research directions in autopilot technology, including Direct Perception, indirect Mediated Perception, and End-to-End Control [1] . this article mainly uses End-to-End Control as the core research method. We use the general neural network and ALexNet, Nvidia-net in CNN as the control side to do training work and evaluate the performance of these three neural networks. We then proposed an improved solution for the Nividia-net algorithm. The flask framework is used to build the server as the control side to analyze the characters of network models and allow the controlled side to send the control information. In order to reduce costs, this research uses a virtual racing game as the controlled side. At the same time, this research collects the information of frontal pictures, steering wheel angles, brakes, throttle, and speed of the racing car through the recording function of the virtual racing game, which is used as training data. Through the above methods, we completed the case of training, running and testing.

A number of different gene selection approaches based on gene expression profiles (GEP) have been developed for tumour classification. A gene selection approach selects the most informative genes from the whole gene space, which is an important process for tumour classification using GEP. This study presents an improved swarm intelligent optimisation algorithm to select genes for maintaining the diversity of the population. The most essential characteristic of the proposed approach is that it can automatically determine the number of the selected genes. On the basis of the gene selection, the authors construct a variety of the tumour classifiers, including the ensemble classifiers. Four gene datasets are used to evaluate the performance of the proposed approach. The experimental results confirm that the proposed classifiers for tumour classification are indeed effective.

The proposed approach does not require any assumption for software failure data.Implementation of the proposed approach is very easy.The proposed SRGMTEF has a reasonable predictive capability. Software reliability growth model (SRGM) with testing-effort function (TEF) is very helpful for software developers and has been widely accepted and applied. However, each SRGM with TEF (SRGMTEF) contains some undetermined parameters. Optimization of these parameters is a necessary task. Generally, these parameters are estimated by the Least Square Estimation (LSE) or the Maximum Likelihood Estimation (MLE). We found that the MLE can be used only when the software failure data to satisfy some assumptions such as to satisfy a certain distribution. However, the software failure data may not satisfy such a distribution. In this paper, we investigate the improvement and application of a swarm intelligent optimization algorithm, namely quantum particle swarm optimization (QPSO) algorithm, to optimize these parameters of SRGMTEF. The performance of the proposed SRGMTEF model with optimized parameters is also compared with other existing models. The experiment results show that the proposed parameter optimization approach using QPSO is very effective and flexible, and the better software reliability growth performance can be obtained based on SRGMTEF on the different software failure datasets.

Display Omitted Knowledge of the samples with and without caption is sufficiently considered.The number of labels is completely determined by the image content.The proposed AIA approach can automatically implemented. Since there is semantic gap between low-level visual features and high-level image semantic, the performance of many existing content-based image annotation algorithms is not satisfactory. In order to bridge the gap and improve the image annotation performance, a novel automatic image annotation (AIA) approach using neighborhood set (NS) based on image distance metric learning (IDML) algorithm is proposed in this paper. According to IDML, we can easily obtain the neighborhood set of each image since obtained image distance can effectively measure the distance between images for AIA task. By introducing NS, the proposed AIA approach can predict all possible labels of the image without caption. The experimental results confirm that the introduction of NS based on IDML can improve the efficiency of AIA approaches and achieve better annotation performance than the existing AIA approaches.

We present a hybrid method using ANN and QPSO for software fault-prone prediction.ANN is used for the classification of software modules.QPSO is controlled more easily than PSO. The identification of a module's fault-proneness is very important for minimizing cost and improving the effectiveness of the software development process. How to obtain the correlation between software metrics and module's fault-proneness has been the focus of much research. This paper presents the application of hybrid artificial neural network (ANN) and Quantum Particle Swarm Optimization (QPSO) in software fault-proneness prediction. ANN is used for classifying software modules into fault-proneness or non fault-proneness categories, and QPSO is applied for reducing dimensionality. The experiment results show that the proposed prediction approach can establish the correlation between software metrics and modules' fault-proneness, and is very simple because its implementation requires neither extra cost nor expert's knowledge. Proposed prediction approach can provide the potential software modules with fault-proneness to software developers, so developers only need to focus on these software modules, which may minimize effort and cost of software maintenance.

BACKGROUND The aim of this study was to analyze the impact and usefulness of characteristic signal change of a linear black signal on magnetic resonance imaging (MRI) on treatment-related decision making. MATERIAL AND METHODS Forty-one patients with a linear black signal on MRI were enrolled in this prospective study. They were randomly divided into the percutaneous kyphoplasty (PKP) group (n=24) and the conservative treatment group (n=17). Clinical measures, including visual analog scale (VAS) and short-form 36 (SF-36) questionnaire, were analyzed. Radiographic measures, including anterior vertebral body height, kyphosis angle and rate of bone-union, were evaluated. RESULTS VAS scores were significantly lower in the PKP group than in the conservative treatment group post-treatment and at one-year follow-up. After one year of treatment, the values for physical functioning, physical health, and body pain were significantly higher in the PKP group than in the conservative treatment group (p

There are serious distortion problems in the history audio and video data. In view of the characteristics of audio data repair, the intelligent technology of audio evaluation is explored. As the traditional audio subjective evaluation method requires a large number of personal to audition and evaluation, the tester’s subjective sense of hearing deviation and sample space data limited the impact of the accuracy of the experiment. Based on the deep learning network, this paper designs an objective quality evaluation system for historical audio and video data and evaluates the performance of the system and the audio signal quality from the perspective of feature extraction and network parameter selection. Experiments show that the system has good performance in this experiment; the predictive results and subjective evaluation of the correlation and dispersion indicators are good, up to 0.91 and 0.19.

Background: HIV long-term non-progressors (LTNP) may hold clues to novel treatment and prevention strategies. We aimed to establish a cohort of HIV LTNP in China and investigate characteristics and determinants of LTNP. Methods: Records in China's national HIV/AIDS information system that met study criteria (i.e., aged =15 years at HIV diagnosis, diagnosis prior to 1 January 2009, and having CD4 count results) were extracted on 30 June 2016. LTNP was defined as diagnosed =8 years with no ART, zero CD4 results

Automatic image annotation (AIA) is a task of assigning one or more semantic concepts to a given image and a promising way to achieve more effective image retrieval and analysis. It is a typical classification problem. Due to the semantic gap between low-level visual features and high-level image semantic, the performances of many existing image annotation algorithms are not satisfactory. This paper presents a novel AIA scheme based on improved quantum particle swarm optimization (IQPSO) algorithm for visual features selection (VFS) and an ensemble stratagem based on boosting technique to improve performance of image annotation approach. To maintain the population diversity, the measure method of population diversity and improvement operation are proposed. To achieve better performance of AIA scheme, the measure of population diversity is as a control condition of VFS process. The classification result of an ensemble classifier is as the final annotation result rather than individual classifier. The experimental results confirm that the proposed AIA scheme is very effectiveness. When using proposed AIA scheme over three image datasets respectively, the annotation results are satisfactory. Display Omitted Measure of population diversity is used as a control condition of feature selection.An improvement operation of QPSO is used to avoid premature convergence.Boosting technique is used for creating an ensemble classifier.

Allergic asthma is characterized by airway hyperresponsiveness, inflammation, and a cellular infiltrate dominated by eosinophils. Numerous epidemiological studies have related the exacerbation of allergic asthma with an increase in ambient inhalable particulate matter from air pollutants. This is because inhalable particles efficiently deliver airborne allergens deep into the airways, where they can aggravate allergic asthma symptoms. However, the cellular mechanisms by which inhalable particulate allergens (pAgs) potentiate asthmatic symptoms remain unknown, in part because most in vivo and in vitro studies exploring the pathogenesis of allergic asthma use soluble allergens (sAgs). Using a mouse model of allergic asthma, we found that, compared with their sAg counterparts, pAgs triggered markedly heightened airway hyperresponsiveness and pulmonary eosinophilia in allergen-sensitized mice. Mast cells (MCs) were implicated in this divergent response, as the differences in airway inflammatory responses provoked by the physical nature of the allergens were attenuated in MC-deficient mice. The pAgs were found to mediate MC-dependent responses by enhancing retention of pAg/IgE/FcεRI complexes within lipid raft–enriched, CD63(+) endocytic compartments, which prolonged IgE/FcεRI-initiated signaling and resulted in heightened cytokine responses. These results reveal how the physical attributes of allergens can co-opt MC endocytic circuitry and signaling responses to aggravate pathological responses of allergic asthma in mice.

Inspired by significant progress in cloud computing technology, a range of applications are increasingly being offered to customers based on cloud servers. In this paper, we design a moving object detection system that enables clients to forward video captured on mobile devices to a server in the cloud. Within the cloud, the video begins to be processed with the moving object in the video detected by the cloud sever. This system enables utilization of the computing ability of the cloud platform and avoids having to rely on the limited processing capacity of mobile devices. As a result, this architecture significantly expands the scope of application for moving object detection.