Search

Deep learning is increasingly being used in high-stake decision making applications that affect individual lives. However, deep learning models might exhibit algorithmic discrimination behaviors with respect to protected groups, potentially posing negative impacts on individuals and society. Therefore, fairness in deep learning has attracted tremendous attention recently. We provide a review covering recent progresses to tackle algorithmic fairness problems of deep learning from the computational perspective. Specifically, we show that interpretability can serve as a useful ingredient to diagnose the reasons that lead to algorithmic discrimination. We also discuss fairness mitigation approaches categorized according to three stages of deep learning life-cycle, aiming to push forward the area of fairness in deep learning and build genuinely fair and reliable deep learning systems.

Small businesses face major challenges to becoming more innovative. These challenges are particularly prevalent in emerging economies where high uncertainties are a barrier to innovation. We know from previous studies that linkages to universities, on the one hand, and public procurement, on the other, support large and innovative firms in their efforts to become more innovative. However, we do not know whether these positive effects also hold true for small businesses. In this paper, we focus on how policy strategies reducing information, market and financial uncertainties shape small businesses’ innovation in China. Based on a sample of 926 small businesses derived from the World Bank Enterprises Survey in China (2012), we find that university-industry linkages enhance innovation, though only when it comes to minor forms of innovation. In line with the resource-based view of the firm, this effect is stronger for small businesses with higher capabilities. Moreover, we show that bidding for or delivering contracts to public sector clients has a positive effect on innovation, and in particular of major forms of innovation. In the bidding selection process, private firms and firms with higher capabilities are selected. Our findings show that both policy strategies have enhanced innovation, though with different effects on the degree of novelty. We attribute this finding to the different degrees of uncertainties they address.

What does it mean for a private enterprise in China to be embedded in a family? Our purpose here is twofold: (1) use social network analysis to describe what it means for a firm to be embedded in a family, (2) reveal from the application a new kind of firm, not family, yet akin to family. Armed with data on a large probability sample of private enterprises — a third of which meet ownership and employment criteria of being family businesses — we uncover a category of “hybrid family firms” that look modern in the style of firms that exclude family, but operate socially in ways similar to family firms. Our conclusion from summary statistics on the sample is that there are no differences in average performance level or network advantage for the three categories of businesses: family firms, hybrid family firms, and family-excluded firms. The fact that CEOs of family firms and hybrid family firms more often turn to family as key business contacts is a fact about network composition that raises no question about network mechanisms. Whether the CEO turns to more or fewer family contacts, government help is more likely with stronger political connections, and business success and survival are more likely with a large, open network. That said, the look-modern, act-traditional hybrid family firms stand alone in prospering with a CEO embedded in a closed business network. Recognition of hybrid family firms adds to the literature’s illustrations of social network analysis used to distinguish types of businesses and business people, and extends the population of organizations within which governance and strategy are likely to be better understood when viewed through a family logic.

Deep learning has shown its great promise in various biomedical image segmentation tasks. Existing models are typically based on U-Net and rely on an encoder-decoder architecture with stacked local operators to aggregate long-range information gradually. However, only using the local operators limits the efficiency and effectiveness. In this work, we propose the non-local U-Nets, which are equipped with flexible global aggregation blocks, for biomedical image segmentation. These blocks can be inserted into U-Net as size-preserving processes, as well as down-sampling and up-sampling layers. We perform thorough experiments on the 3D multimodality isointense infant brain MR image segmentation task to evaluate the non-local U-Nets. Results show that our proposed models achieve top performances with fewer parameters and faster computation.

This paper reports a 3D printing based manufacturing method using biomass-fungi composite material. The biomass is derived from waste agricultural materials. The fungi grow through the biomass and bind the biomass together. The novel method comprised the preparation of a printable biomass-fungi mixture. After printing, the fungi grew inside the printed sample over a few days. The feasibility of this new method was demonstrated by a preliminary experiment. In comparison to molding based manufacturing processes, 3D printing can facilitate greater flexibility in design and print custom designs for packaging and construction applications.

It is known that 3D printing can facilitate greater design flexibility in the printing of custom shapes for packaging and construction applications using biomass–fungi composite materials. The feasibility of this new method was demonstrated by a preliminary experiment, the results of which were reported in a journal publication in 2020. As a follow-up, this paper reports on an experimental study on the relationship between the mixture composition (i.e., the psyllium husk powder content) and print quality using this new method. Four mixtures were prepared by varying the amounts of psyllium husk powder (in grams) added to 400 mL of water. The ratios (g/mL) of psyllium husk powder weight (wp) over volume of water (vw) for the mixtures were 0, 1:40, 2:40, and 3:40. Each mixture also contained 100 g of biomass–fungi material and 40 g of whole wheat flour. The print quality of the samples was evaluated based on the extrudability and shape stability. The results showed that mixtures without any psyllium husk powder were not extrudable. An increase in the ratio of psyllium husk powder to water from 1:40 to 2:40 resulted in improved print quality, however, when the psyllium husk powder to water ratio was increased to 3:40, the extrudability became worse. This phenomenon was explained by analyzing the rheological properties of the mixtures.

BackgroundHydatidiform mole is a disease caused by abnormal proliferation of trophoblastic cells. MiR-30a acts as a tumor suppressor gene in most tumors and participates in the development of various cancers. But its role in hydatidiform mole is not clear. MethodsRT-qPCR was used to verify the expression level of miR-190 and STOX2. Flow cytometry (FCM) assays were performed to detect cell cycle. CCK-8 assay, EDU assay and colony formation assay was used to detect proliferation ability. Transwell assay was used to test invasion ability. Dual-luciferase reporter assay and Western blotting were used to elucidate the potential mechanisms involved.ResultThe low expression of miR-30a promoted the ability of proliferation, migration and invasion in trophoblastic cells (JAR and HTR-8). The dual luciferase assay confirmed that STOX2 was a new target of miR-30a and resisted the effect of up-regulated miR-30a in trophoblastic cells. In addition, up-regulation of STOX2 by miR-30a could activate ERK, AKT and P38 signaling pathways. These results revealed that a new mechanism for ERK, AKT and P38 activation by miR-30a/STOX2, which could create excessive proliferation of trophoblast cells in the hydatidiform mole.Conclusions In this study, we found that miR-30a plays an important role in the development of HM. Taken together, these fndings indicate that miR-30a may effect the malignant transformation of human trophoblastic cells by downregulating STOX2, which strengthens our understanding of miR-30a in regulating BC cell transformation.

Image captioning has made substantial progress with huge supporting image collections sourced from the web. However, recent studies have pointed out that captioning datasets, such as COCO, contain gender bias found in web corpora. As a result, learning models could heavily rely on the learned priors and image context for gender identification, leading to incorrect or even offensive errors. To encourage models to learn correct gender features, we reorganize the COCO dataset and present two new splits COCO-GB V1 and V2 datasets where the train and test sets have different gender-context joint distribution. Models relying on contextual cues will suffer from huge gender prediction errors on the anti-stereotypical test data. Benchmarking experiments reveal that most captioning models learn gender bias, leading to high gender prediction errors, especially for women. To alleviate the unwanted bias, we propose a new Guided Attention Image Captioning model (GAIC) which provides self-guidance on visual attention to encourage the model to capture correct gender visual evidence. Experimental results validate that GAIC can significantly reduce gender prediction errors with a competitive caption quality. Our codes and the designed benchmark datasets are available at https://github.com/datamllab/Mitigating_Gender_Bias_In_Captioning_System.

Back propagation based visualizations have been proposed to interpret deep neural networks (DNNs), some of which produce interpretations with good visual quality. However, there exist doubts about whether these intuitive visualizations are related to the network decisions. Recent studies have confirmed this suspicion by verifying that almost all these modified back-propagation visualizations are not faithful to the model's decision-making process. Besides, these visualizations produce vague "relative importance scores", among which low values can't guarantee to be independent of the final prediction. Hence, it's highly desirable to develop a novel back-propagation framework that guarantees theoretical faithfulness and produces a quantitative attribution score with a clear understanding. To achieve the goal, we resort to mutual information theory to generate the interpretations, studying how much information of output is encoded in each input neuron. The basic idea is to learn a source signal by back-propagation such that the mutual information between input and output should be as much as possible preserved in the mutual information between input and the source signal. In addition, we propose a Mutual Information Preserving Inverse Network, termed MIP-IN, in which the parameters of each layer are recursively trained to learn how to invert. During the inversion, forward Relu operation is adopted to adapt the general interpretations to the specific input. We then empirically demonstrate that the inverted source signal satisfies completeness and minimality property, which are crucial for a faithful interpretation. Furthermore, the empirical study validates the effectiveness of interpretations generated by MIP-IN.

Proper expression of the genes plays a vital role in the function of an organism. Recent advancements in DNA microarray technology allow for monitoring the expression level of thousands of genes. One of the important tasks in this context is to understand the underlying mechanisms of gene regulation. Recently, researchers have focused on identifying local DNA elements, or motifs to infer the relation between the expression and the nucleotide sequence of the gene. This study proposes a novel data adaptive representation approach for supervised learning to predict the response associated with the biological sequences. Biological sequences such as DNA and protein are a class of categorical sequences. In machine learning, categorical sequences are generally mapped to a lower dimensional representation for learning tasks to avoid problems with high dimensionality. The proposed method, namely SW-RF (sliding window-random forest), is a feature-based approach requiring two main steps to learn a representation for categorical sequences. In the first step, each sequence is represented by overlapping subsequences of constant length. Then a tree-based learner on this representation is trained to obtain a bag-of-words like representation which is the frequency of subsequences on the terminal nodes of the tree for each sequence. After representation learning, any classifier can be trained on the learned representation. A lasso logistic regression is trained on the learned representation to facilitate the identification of important patterns for the classification task. Our experiments show that proposed approach provides significantly better results in terms of accuracy on both synthetic data and DNA promoter sequence data. Moreover, a common problem for microarray datasets, namely missing values, is handled efficiently by the tree learners in SW-RF. Although the focus of this paper is on biological sequences, SW-RF is flexible in handling any categorical sequence data from different applications.

Outlier detection is an important task for various data mining applications. Current outlier detection techniques are often manually designed for specific domains, requiring large human efforts of database setup, algorithm selection, and hyper-parameter tuning. To fill this gap, we present PyODDS, an automated end-to-end Python system for Outlier Detection with Database Support, which automatically optimizes an outlier detection pipeline for a new data source at hand. Specifically, we define the search space in the outlier detection pipeline, and produce a search strategy within the given search space. PyODDS enables end-to-end executions based on an Apache Spark backend server and a light-weight database. It also provides unified interfaces and visualizations for users with or without data science or machine learning background. In particular, we demonstrate PyODDS on several real-world datasets, with quantification analysis and visualization results., Comment: In Companion Proceedings of the Web Conference 2020 (WWW 20)

This study examines whether and how anticorruption efforts may mitigate the risk of corporate fraud. Based on a sample of Chinese publicly listed firms over the period of 2008 to 2017, we find that anticorruption efforts reduce the likelihood of fraud commission and increase the likelihood of detection given fraud. These effects are driven by state-owned enterprises and politically connected firms through politician board members. We also find that firms located in regions with well-developed market and legal institutions are less likely to commit fraud in the post anticorruption period. Firms increasing internal monitoring by appointing local independent directors with accounting background help to explain the reduction of fraud in these regions. This study contributes to the literature on corporate wrongdoing and the design of strategies to mitigate the risk of corporate fraud in an emerging economy context.

What does it mean for a private enterprise in China to be embedded in a family? Our purpose here is twofold: (1) use social network analysis to describe what it means for a firm to be embedded in a family, (2) reveal from the application a new kind of firm, not family, yet akin to family. Armed with data on a large probability sample of private enterprises — a third of which meet ownership and employment criteria of being family businesses — we uncover a category of “hybrid family firms” that look modern in the style of firms that exclude family, but operate socially in ways similar to family firms. Our conclusion from summary statistics on the sample is that there are no differences in average performance level or network advantage for the three categories of businesses: family firms, hybrid family firms, and family-excluded firms. The fact that CEOs of family firms and hybrid family firms more often turn to family as key business contacts is a fact about network composition that raises no question about network mechanisms. Whether the CEO turns to more or fewer family contacts, government help is more likely with stronger political connections, and business success and survival are more likely with a large, open network. That said, the look-modern, act-traditional hybrid family firms stand alone in prospering with a CEO embedded in a closed business network. Recognition of hybrid family firms adds to the literature’s illustrations of social network analysis used to distinguish types of businesses and business people, and extends the population of organizations within which governance and strategy are likely to be better understood when viewed through a family logic.

Neuron tracing, also known as neuron reconstruction, studies 3D morphologies of neurons based on imaging data. Neuron reconstruction is of fundamental importance in computational neuroscience since it is a crucial step towards reverse engineering of the wiring and functions of a brain. On the other hand, it is not possible to manually trace all neurons due to the complexity and cost of this task. Hence, it raises the need of building a computational pipeline to perform automatic neuron reconstruction. In this work, we propose a deep learning approach for improving the accuracy of 3D neuron reconstruction. First, we propose to learn shared features among different images in the whole dataset. Such shared features are learned automatically by our model at different scales. Second, we propose to incorporate such features to guide the information flow in the network. Specifically, we propose to build skip connections between the encoder and the decoder of our networks by incorporating the hierarchical and shared features. Our proposed skip connections are built based on the attention mechanism, where the hierarchical shared features serve as the query matrix and the local input features serve as the key and value matrices. Since the parameters are learned automatically, we expect that only useful spatial information is transmitted to the decoder. We conduct both qualitative and quantitative experiments to demonstrate the effectiveness of our proposed method. Experimental results show that our proposed model has the ability to capture detailed structural information for neurons. Our results also demonstrate that the proposed model is robust to noise. In addition, quantitative evaluations show that our method achieves better performance than other approaches.

Parkinson’s disease (PD) is a common neurodegenerative disorder disease. Identifying an accurate model to predict severity level is critical to prevent severe suffering for PD patients. However, the existing research does not consider the heterogeneity among patients, falls short for prediction uncertainty characterization and typically adopts models that involve tedious parameter tuning processes. We propose to incorporate transfer learning and sparse learning under a Hierarchical Bayesian framework for tractable estimation of parameter posterior distribution and prediction uncertainty quantification. Specifically, we develop an empirical Bayes transfer learning (ebTL) model that accounts for patient heterogeneity and meanwhile allows for knowledge transfer between the modeling processes of different patients. ebTL is also featured for automatic hyper-parameters estimation without a tedious tuning process. Finally, we present an application of predicting PD severity level by using features extrac...

Additive manufacturing (AM) has had an enormous impact in the manufacturing sector. Its role has evolved from printing prototypes to manufacturing functional parts for a variety of applications in the automotive, aerospace, and medical industries. Recently, AM processes have also been applied in the infrastructure construction industry. Applications of AM processes could bring in significant improvements in infrastructure construction, specifically in the areas of productivity and safety. It is desirable to have a review on the current state of emerging AM processes for infrastructure construction as well as existing gaps in this field. This paper reviews AM processes in infrastructure construction. It discusses the process principle, application examples, and gaps for each of the AM processes.

3D printing processes have found several applications in automobile, aerospace, and biomedical industries. Recent development of large-scale 3D printing processes has fueled applications in the construction industry. Materials such as ceramic clay (porcelain) can now be used in additive manufacturing (AM). However, information regarding the effect of process parameters on part quality is limited. There is a need to study the effects of various parameters on the successful printing of samples for the construction industry. This paper presents an experimental study to determine the feasible parameter regions for extrusion-based 3D printing of porcelain using a Delta WASP 2040 Clay printer. The printing parameters studied are extruder height, layer thickness, print speed and air pressure. The effects of these parameters on the surface quality of the printed samples are examined. The identified feasible parameter regions from this study would provide preliminary insights on material extrusion-based 3D printing of porcelain.

Anomaly detection aims to distinguish observations that are rare and different from the majority. While most existing algorithms assume that instances are i.i.d., in many practical scenarios, links describing instance-to-instance dependencies and interactions are available. Such systems are called attributed networks. Anomaly detection in attributed networks has various applications such as monitoring suspicious accounts in social media and financial fraud in transaction networks. However, it remains a challenging task since the definition of anomaly becomes more complicated and topological structures are heterogeneous with nodal attributes. In this paper, we propose a spectral convolution and deconvolution based framework -- SpecAE, to project the attributed network into a tailored space to detect global and community anomalies. SpecAE leverages Laplacian sharpening to amplify the distances between representations of anomalies and the ones of the majority. The learned representations along with reconstruction errors are combined with a density estimation model to perform the detection. They are trained jointly as an end-to-end framework. Experiments on real-world datasets demonstrate the effectiveness of SpecAE., Comment: 7 pages, in proceedings of the 28th ACM International Conference on Information and Knowledge Management (CIKM)

Social networks play a crucial role in entrepreneurial processes and outcomes. However, little is known about the impact of social networks on microenterprises’ business performance. Addressing this gap, I focus on two dimensions of social networks, namely network size and resources embedded in a social network. Theoretically, combining insights from strategy research highlighting strategic needs as a determinant for external resource acquisition and the resource-based view of the firm suggesting that acquiring complementary resources can be a source of competitive advantage, I theorize why micro-entrepreneurs need to acquire external resources from their social networks, and why and how acquiring external resources are beneficial for their business performance. Empirically, I investigate the social network-performance relationship in the microenterprises context based on a dataset of 200 stratified randomly sampled Chinese micro-entrepreneurs in Shanghai. While the results suggest that micro-entrepreneurs’ network size per se does not relate to their business performance, I show that micro-entrepreneurs who can access to more varieties of resources embedded in their social networks tend to report a higher profit in a typical month. Furthermore, I find that micro-entrepreneurs whose external resources are more equally distributed across network contacts are associated with better business performance.

In the process of interactive design generally through the interior design graphics to achieve customer needs, interactive design is a new type of green design system, at present, has been widely used in computer web production, industrial product design, furniture manufacturing and other fields. With the continuous development of computer technology, the human-computer interaction problem is gradually reflected. In order to effectively solve this problem, we add the concept of interaction design into the field of architectural design. The interactive design is combined with the characteristics of architectural design to analyze the architectural design. In this paper, the composition, technology platform and operation mode of the interactive design system are analyzed and studied, and the residential building technology is discussed from these three aspects. In addition, it is applied to the actual case, through the use of digital design, numerical control processing simulation experiment, image experiment and other scientific research methods, the residential building space environment design experimental study. This will also lead to reference standards for green interior design of public and industrial buildings.

Recently, long-range surface plasmon resonance (LRSPR) sensor has attracted a great deal of attention as a potentially non-destructive and label-free technique for cellular studies in real time. Thus, much effort has been placed on the fabrication and optimization of multilayered structure required for the excitation of LRSPR. In this work, a detailed study about the influence of both plasma polymerized dielectric buffer layer (DBL) and thin gold film on the excitation of LRSPR was performed. The DBLs of different thicknesses were deposited directly onto SF11 glass slides by radio frequency plasma polymerization (pp) of perfluorooctyl ethylene (PFOE). Thereafter, Au films of different thicknesses were thermally evaporated onto the ppPFOE layers. Atomic force microscopy (AFM) results suggest that the resulting SF11/ppPFOE/Au structure has a smooth surface regardless of Au film’s thickness. LRSPR measurements indicate that the excitation of LRSPR relies not only on the thickness of the ppPFOE buffer layer, but also on the thickness and optical property of thin Au film. Theoretical simulation based on Fresnel’s equation allows for the determination of both the thickness and optical constant of each layer supporting the LRSPR, and also enables us to predict the optimum combination of ppPFOE and Au film in a LRSPR sensor. The performance of various LRSPR sensors to monitor the bulk refractive index variation has also been investigated.

Sanctions against bad behavior vary greatly. Some wrongdoers are regarded as no longer trustworthy; others face only modest consequences. This study investigates how this difference can be explaine...

Additive manufacturing (AM) has applications in several fields ranging from aerospace and consumer goods to the medical industry. However, applications of AM in civil infrastructure design and construction are very limited. Based on information shared at the NSF workshop on Additive Manufacturing (3D Printing) for Civil Infrastructure Design and Construction in July 2017, this paper summarizes the current state of the field, gaps, and recommendations.

With the constant popularization of computers in recent years, people have already entered an era of network, and it is not difficult to discover that people will achieve information about current affairs, recreation and entertainment from the internet. Consequently, IP network enjoys unprecedented development. However, some unavoidable security problems shall be solved during the development process.

Bing, L.; Xing, Q.-G.; Liu, X., and Zou, N.-N., 2019. Spatial distribution characteristics of oil spills in the Bohai Sea based on satellite remote sensing and GIS. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Remote Sensing and Geoscience Information Systems of Coastal Environments. Journal of Coastal Research, Special Issue No. 90, pp. 164-170. Coconut Creek (Florida), ISSN 0749-0208.In recent years, on account of the vast and persistent damage of oil spill accidents, it becomes essential to carry out a further study on oil spill distribution characteristics on sea. Routine satellite remote sensing surveillance on oil spills with Synthetic Aperture Radar (SAR) proved to be ideal for analyzing distribution of oil pollution in macro scale. In this research work, considering the presence of “look-alikes” phenomena on SAR images as well as current operational application of “confidence level”, a confidence-oriented oil spill geodatabase is initially designed and built for quantification analysis. Then, in view of the requirement of marine grid management, a common framework based on remote sensing and Geographic Information System (GIS) is proposed to map and reveal the spatial distribution of oil pollution, in the process of which, oil pollution index (OPI) is put forward to evaluate oil pollution levels, then oil pollution distribution map can be compiled in terms of OPI, and the relationship between oil pollution and oil spill risk source can be further analyzed. Finally, a case study of the Bohai Sea and the north of the Yellow Sea with five-year's inter-annual satellite monitoring data was studied. The results showed that, during the period between 2009 and 2013, the high frequency grids of potential oil spills mainly distributed in the Bohai Bay and the south of the Liaodong Bay. 77.03 % of detected oil spills were within a certain distance along the sea lanes or near the offshore platforms, indicating a high risk of oil pollution of these areas. Generally, this research work also benefits for yield for the first time a rather comprehensive distribution characteristics of potential oil spills in macro scale in the Bohai Sea combined with remote sensing and GIS technique, which will be beneficial for oil spill response, preparedness and risk management.

Dynamic network data are often encountered in social, biological, and engineering domains. There are two types of variability in dynamic network data: variability of natural evolution and variability due to assignable causes. The latter is the “change” referred to in this article. Accurate and timely change detection from dynamic network data is important. However, it has been infrequently studied, with most of the existing research having focused on community detection, prediction, and visualization. Change detection is a classic research area in Statistical Process Control (SPC), and various approaches have been developed for dynamic data in the form of univariate or multivariate time series but not in the form of networks. We propose a Network State Space Model (NSSM) to characterize the natural evolution of dynamic networks. For tractable parameter estimation of the NSSM, we develop an Expectation Propagation algorithm to produce an approximation for the observation equation of the NSSM and then use Expectation–Maximization integrated with Bayesian Optimal Smoothing to estimate the parameters. For change detection, we further propose a Singular Value Decomposition (SVD)-based method that integrates the NSSM with SPC. A real-world application on Enron dynamic email networks is presented, in which our method successfully detects two known changes.

The dynamic testing of stress wave reflection method is an easy, economical, rapid and reliable nondestructive testing technology and has been perfected used in the pile foundation dynamic testing. Anchor rod is more in line with the wave theory of one-dimensional elastic rod than pile foundation whether to theory or the geometric solid, material, stress state and other factors. The quality of anchor rod is determined by using small strain method (hammering method) of pile foundation dynamic measurement. Four testing anchor rods with different length and integrity were made in project of "China Marine Sport School", stress wave time domain curve of anchor rod was obtained by ZK-7E portable intelligent testing of piles instrument, defect location, length of anchor rod and integrity reflected by measured waveform are consistent with the actual situation, the method using pile foundation dynamic testing technology for anchor rod nondestructive testing is proved feasible and reliable.

Background Serum cystatin C (Cys-C), an inhibitor of cysteine proteases, has been suggested as an ideal biomarker of glomerular filtration rate (GFR). Objectives The objective of this study was to describe the reference intervals of serum Cys-C and identify factors associated with serum Cys-C or its variability, including age, gender, creatinine (Crea), blood urea nitrogen (BUN), and uric acid (UA). Design and Methods Serum Cys-C, Crea, BUN, and UA were measured in 4,517 healthy participants aged 8–89 years attending our hospital. Serum Cys-C was analyzed using a latex-enhanced immunoturbidimetric method. Crea were tested by picric acid jaffe method, BUN, and UA by kinetic UV assays. Results The predominant characteristic of Cys-C distribution was that Cys-C concentration in age ≥60 years group was the highest (P < 0.05). The differences of Cys-C concentration between males and females existed for subjects aged from 30 to 59 years (P < 0.05). In a multiple model adjusted only for gender and age, gender (β = 0.007) has stronger effect on Cys-C levels, compared with age (β = 0.003). The clinical variables, comprised of age, gender, Crea, BUN, and UA, involved in the fully adjusted equation accounted for 37.6% of variation of Cys-C. Conclusions Ninety-five percent reference intervals for healthy population were partitioned into three categories only by age, 0.59–1.07 mg/L for subjects aged 19–59 years; 0.74–1.14 mg/L for the older aged ≥60 years; and 0.63–1.11 mg/L for children aged ≤18 years. Serum Cys-C is significantly related to gender, age, UA, Crea, and BUN. Besides, there are still other factors contributing to variation of Cys-C levels.

A finite element model is developed in this paper for simulating cyclic deformation in ultra-fined polycrystalline metals. The grain material is characterized by conventional theory of mechanism-based strain gradient plasticity (CMSG). A cohesive interface model was used to simulate the initiation and propagation of intergranular cracks. The simulation results show that inhomogeneous plastic deformation induces high strain gradient effects and severely plastic hardening in the grain interior, and the intergranular crack has a significant influence on the overall mechanical properties of ultra-fined polycrystalline metals subjected to cyclic loading.

This paper contributes to the ongoing discussion about the supply chain management (SCM) of Chinese toy makers by conducting a survey among 148 respondents from the industry. The questionnaire includes 24 questions that belong to six sections, which are design and buying, materials procurement, production operations, product distribution, quality control, list of tools (technology). Through analyzing the feedback data from 36 respondents, this study reveals the update circumstances of Chinese toy industry regarding SCM. Notably, on the one hand, most toy makers (about 90%) suggest that they value quality control, which is essential for global competition. On the other hand, less toy makers prefer to adopt advanced SCM tools; this may hinder their competitiveness in the long run. To conclude, both constructive and practical suggestions are given to the toy makers for improving their SCM performance.

Although arachnoid granulations (AGs) were initially described by Pacchioni more than 300 years ago, they are still poorly described, especially those in middle cranial fossa. The aim of this study was to identify anatomical features of AGs in middle cranial fossa of cadaver and compare such features with that of 64-slice computed tomography.The study involved anatomical dissection of 33 adult cadaveric heads and computed tomographic (CT) examinations of 40 patients with various intracranial diseases. The number, size, distribution, and morphology of the AGs in middle cranial fossa of the cadavers and patients were examined and compared.The number of AGs observed on the middle cranial fossa of the cadaveric specimens was greater than that of the patients (P0.05). While anatomic dissection revealed a total of 228 AGs in 24 of 33 cadaveric heads, CT scans demonstrated only 80 AGs in 23 of 40 patients. In cadavers, the AGs occurred in the following sites in order of frequency: the middle meningeal sinus (144 AGs, 63.2%), sphenoparietal sinus (47 AGs, 20.6%), lateral foramen rotundum (19 AGs, 8.3%), and cavernous sinus (18 AGs, 7.9%). In CT images, the AGs occurred in the following sites in order of frequency: the middle meningeal sinus (50 AGs, 62.5%), sphenoparietal sinus (23 AGs, 28.8%), and lateral foramen rotundum (7 AGs, 8.8%). The AGs of cavernous sinus and venous lacunae adjacent to the middle meningeal sinus were hardly identified on CT images. Most of the AGs were spherical or finger-like in shape. Histologically, the AGs can be divided into two types: single type and lobulated type.The study provides a detailed description of AGs in or near the middle meningeal sinus, sphenoparietal sinus, lateral foramen rotundum, and cavernous sinus. It also reveals a difference in the ability of detecting cranial AGs between microanatomy and CT scans.