Search

Dendrite core is the center point of the dendrite. The information of dendrite core is very helpful for material scientists to analyze the properties of materials. Therefore, detecting the dendrite core is a very important task in the material science field. Meanwhile, because of some special properties of the dendrites, this task is also very challenging. Different from the typical detection problems in the computer vision field, detecting the dendrite core aims to detect a single point location instead of the bounding-box. As a result, the existing regressing bounding-box based detection methods can not work well on this task because the calculated center point location based on the upper-left and lower-right corners of the bounding-box is usually not precise. In this work, we formulate the dendrite core detection problem as a segmentation task and proposed a novel detection method to detect the dendrite core directly. Our whole pipeline contains three steps: Easy Sample Detection (ESD), Hard Sample Detection(HSD), and Hard Sample Refinement (HSR). Specifically, ESD and HSD focus on the easy samples and hard samples of dendrite cores respectively. Both of them employ the same Central Point Detection Network (CPDN) but do not share parameters. To make HSD only focus on the feature of hard samples of dendrite cores, we destroy the structure of the easy samples of dendrites which are detected by ESD and force HSD to learn the feature of hard samples. HSR is a binary classifier which is used to filter out the false positive prediction of HSD. We evaluate our method on the dendrite dataset. Our method outperforms the state-of-the-art baselines on three metrics, i.e., Recall, Precision, and F-score.

Shadow removal is to restore shadow regions to their shadow-free counterparts while leaving non-shadow regions unchanged. State-of-the-art shadow removal methods train deep neural networks on collected shadow & shadow-free image pairs, which are desired to complete two distinct tasks via shared weights, i.e., data restoration for shadow regions and identical mapping for non-shadow regions. We find that these two tasks exhibit poor compatibility, and using shared weights for these two tasks could lead to the model being optimized towards only one task instead of both during the training process. Note that such a key issue is not identified by existing deep learning-based shadow removal methods. To address this problem, we propose to handle these two tasks separately and leverage the identical mapping results to guide the shadow restoration in an iterative manner. Specifically, our method consists of three components: an identical mapping branch (IMB) for non-shadow regions processing, an iterative de-shadow branch (IDB) for shadow regions restoration based on identical results, and a smart aggregation block (SAB). The IMB aims to reconstruct an image that is identical to the input one, which can benefit the restoration of the non-shadow regions without explicitly distinguishing between shadow and non-shadow regions. Utilizing the multi-scale features extracted by the IMB, the IDB can effectively transfer information from non-shadow regions to shadow regions progressively, facilitating the process of shadow removal. The SAB is designed to adaptive integrate features from both IMB and IDB. Moreover, it generates a finely tuned soft shadow mask that guides the process of removing shadows. Extensive experiments demonstrate our method outperforms all the state-of-the-art shadow removal approaches on the widely used shadow removal datasets.

In developing countries, anti-poverty programs are often implemented by local governments. However, due to the limitation of fiscal resources, the amount of anti-poor expenditure by the local government is generally less than what is needed for the poor. In this paper, we investigate whether an increase in the fiscal resources of local government will lead to an increase in anti-poor fiscal expenditure using county-level Chinese data. Using the fixed effect model, we show that local governments will put more fiscal resources into the minimum living standard guarantee (MLSG) system if they receive more intergovernmental transfers from high-level governments, but this effect only exists in urban areas. Moreover, the off-budget fiscal revenue does not affect the anti-poverty expenditure, both in rural and urban areas.

Background This study aimed to retrospectively explore the safety and feasibility of computed tomography (CT)‐guided synchronous microwave ablation (MWA) followed by core‐needle biopsy (CNB) via a coaxial cannula for highly suspected malignant lung ground‐glass opacities (GGOs). Methods The clinical data of 66 patients (66 GGOs) treated with CT‐guided synchronous MWA followed by CNB via a coaxial cannula from January 2019 to January 2021 were included in this study. The technical success rate, curative effect, and complications were evaluated. Results Technical success rates were 100%. The pneumothorax rate was 36.4% (24/66). 72.7% (48/66) patients had the bronchopulmonary hemorrhage, 81.3% of hemorrhage was attributable to CNB. 24.2% (16/66) patients had varying degrees of pleural effusion. The pathological results were adenocarcinomas (n = 44), atypical adenomatous hyperplasia (n = 2), chronic inflammation (n = 3) and indeterminate pathological diagnosis (n = 17) with a 69.7% (46/66) positive diagnosis rate. The therapeutic response rate was 100.0% (66/66). Conclusions Synchronous MWA followed by CNB via a coaxial cannula has a satisfactory ablation effectiveness and an acceptable biopsy positive rate, which is an alternative treatment for highly suspected malignant GGOs., Synchronous microwave ablation followed by core‐needle biopsy can reduce the influence of biopsy‐induced pneumothorax and pulmonary hemorrhage on microwave ablation, which has a satisfactory therapeutic effect and an acceptable diagnostic positivity rate.

Objective To explore the diagnostic values of different post-processing modes of dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) in the differentiation of glioma grade. Methods A total of 89 patients with glioma confirmed by postoperative pathology in our hospital from January 2013 to July 2019 were enrolled, including 29 patients of gradeⅡ, 24 of grade Ⅲ and 36 of grade Ⅳ. All patients underwent preoperative DCE-MRI examination, and corresponding parameters were collected and processed using post-processing software Tissue-4D and Omni-Kinetics (OK) respectively. One-way ANOVA was adopted to compare the differences of the obtained parameters among glioma of different grades. Spearman rank correlation analysis was performed to analyze the correlation between each parameter and glioma grade, and receiver operating characteristic (ROC) curve was used to evaluate the area under the curve (AUC) of each parameter. Results In hot spot analysis, the Ktrans values of both post-processing modes, as well as the iAUC value of Tissue-4D and the Vp value of OK were all significantly different among glioma of different grades (P < 0.05). The Kep value of grade Ⅳ glioma was obviously lower than that of grade Ⅱ glioma (P < 0.05), while the Ve values of grade Ⅳ and Ⅲ tumors were notably higher than that of gradeⅡ one (P < 0.05). The AUC value of Ktrans (Tissue-4D) was 0.949 in the differentiation of gradeⅡ from grade Ⅳ glioma (P < 0.01), indicating the highest diagnostic efficiency, and the correlation between Ktrans (OK) and glioma grade was the highest (r=0.775, P < 0.01). In the histogram analysis of OK, the 50th percentile of Ktrans value, the mean and the 75th and 90th percentiles of Ve value, the mean, entropy and the 50th percentile of Kep value, as well as the mean value of Vp showed significances in the differentiation of grade Ⅱ from grade Ⅲ and grade Ⅱ from grade Ⅳ glioma (P < 0.05). Moreover, the energy, entropy and 50th, 75th, 90th percentiles of Vp value, and the energy of Kep value were statistically different in the identification of grade Ⅳ from grade Ⅱ and grade Ⅳ from grade Ⅲ (P < 0.05). Conclusion The Ktrans and Ve values of the 2 post-processing modes have rather high diagnostic efficiency for glioma grading in the hotspot analysis, and the histogram analysis of OK software involves more abundant and comprehensive parameters.

Fully-supervised shadow removal methods achieve the best restoration qualities on public datasets but still generate some shadow remnants. One of the reasons is the lack of large-scale shadow & shadow-free image pairs. Unsupervised methods can alleviate the issue but their restoration qualities are much lower than those of fully-supervised methods. In this work, we find that pretraining shadow removal networks on the image inpainting dataset can reduce the shadow remnants significantly: a naive encoder-decoder network gets competitive restoration quality w.r.t. the state-of-the-art methods via only 10% shadow & shadow-free image pairs. After analyzing networks with/without inpainting pre-training via the information stored in the weight (IIW), we find that inpainting pretraining improves restoration quality in non-shadow regions and enhances the generalization ability of networks significantly. Additionally, shadow removal fine-tuning enables networks to fill in the details of shadow regions. Inspired by these observations we formulate shadow removal as an adaptive fusion task that takes advantage of both shadow removal and image inpainting. Specifically, we develop an adaptive fusion network consisting of two encoders, an adaptive fusion block, and a decoder. The two encoders are responsible for extracting the feature from the shadow image and the shadow-masked image respectively. The adaptive fusion block is responsible for combining these features in an adaptive manner. Finally, the decoder converts the adaptive fused features to the desired shadow-free result. The extensive experiments show that our method empowered with inpainting outperforms all state-of-the-art methods.

Disentangled representation learning remains challenging as ground truth factors of variation do not naturally exist. To address this, we present Vocabulary Disentanglement Retrieval~(VDR), a simple yet effective retrieval-based disentanglement framework that leverages nature language as distant supervision. Our approach is built upon the widely-used bi-encoder architecture with disentanglement heads and is trained on data-text pairs that are readily available on the web or in existing datasets. This makes our approach task- and modality-agnostic with potential for a wide range of downstream applications. We conduct experiments on 16 datasets in both text-to-text and cross-modal scenarios and evaluate VDR in a zero-shot setting. With the incorporation of disentanglement heads and a minor increase in parameters, VDR achieves significant improvements over the base retriever it is built upon, with a 9% higher on NDCG@10 scores in zero-shot text-to-text retrieval and an average of 13% higher recall in cross-modal retrieval. In comparison to other baselines, VDR outperforms them in most tasks, while also improving explainability and efficiency.

This study was designed to investigate the potential effects of chronic stress on periodontal bone remodeling and its mechanism during orthodontic tooth movement in rats. Forty-eight male SD rats aged 8 weeks were randomly divided into control group and chronic stress group. Chronic unpredictable mild stress model (CUMS) was established in the stress group, which was validated by behavioral experiment as well as cortisol (CORT) and adrenocorticotropic hormone (ATCH) levels. Then, the two groups were further divided into three distinct groups, namely group with no orthodontic force, group with 30 g orthodontic force and group with 50 g orthodontic force respectively to construct orthodontic tooth movement model. The rats were sacrificed after 14 days and maxilla on the loading side was obtained to measure tooth movement distance. It was found that compared with the control group, the chronic stress group displayed increased parathyroid hormone (PTH), amino terminal peptide of type I procollagen (PINP) and c-terminal peptide of type I collagen branch (CTX) levels as measured by enzyme-linked immunosorbent assay (ELISA). Hematoxylin and Eosin (HE) and TRAP staining showed fewer osteoblasts and more number of osteoclasts. The results of western blot showed no significant change in expression of Adenylate cyclase (AC) but increased phospholipase C (PLC) levels were noted. In addition, increased NF-κB expression was observed by immunohistochemistry. Overall, chronic stress can affect bone remodeling during orthodontic tooth movement by increasing the content of PTH in the blood and increasing PLC and NF-κB.

This study aimed to analyse changes in the nasal cavity and maxillary sinus structure and function in patients with skeletal class III malocclusion 1 year after bimaxillary surgery.In this study, cone-beam computed tomography (CBCT) images of 20 patients (10 men and 10 women; mean age 24.3 ± 3.4 years) with skeletal class III malocclusion who underwent Le Fort I osteotomy and bilateral sagittal split osteotomy were obtained before and 1 year after the surgery. CBCT data were stored opened with element 3D (E3D) to establish a nasal airway model (the paranasal sinus includes only the maxillary sinus). Ansys (ANSYS) software is used for simulation and analysis.The maxillary sinus and nasal cavity volumes decreased significantly 1 year after the surgery. After surgery, the volume of nasal cavity decreased by 13.5%, and the average volume of maxillary sinus decreased by 7.8%. There was no significant difference in the degree of deviation of the septum and nasal cavity resistance, and air distribution in the maxillary sinus did not change. The nasal cavity wall shear stress change was similar to that before surgery.The maxillary sinus volume and nasal cavity volume of patients with skeletal class III malocclusion changed significantly after bimaxillary surgery, but there was no significant change in nasal ventilation function 1 year after surgery.

Twisted stacking of van der Waals materials with moiré superlattices offers a new way to tailor their physical properties via engineering of the crystal symmetry. Unlike well-studied twisted bilayers, little is known about the overall symmetry and symmetry-driven physical properties of continuously supertwisted multilayer structures. Here, using polarization-resolved second harmonic generation (SHG) microscopy, we report threefold (C3) rotational symmetry breaking in supertwisted WS2 spirals grown on non-Euclidean surfaces, contrasting the intact symmetry of individual monolayers. This symmetry breaking is attributed to a geometrical magnifying effect in which small relative strain between adjacent twisted layers (heterostrain), verified by Raman spectroscopy and multiphysics simulations, generates significant distortion in the moiré pattern. Density-functional theory calculations can explain the C3 symmetry breaking and unusual SHG response by the interlayer wave function coupling. These findings thus pave the way for further developments in the so-called "3D twistronics".

Previous works show that Pre-trained Language Models (PLMs) can capture factual knowledge. However, some analyses reveal that PLMs fail to perform it robustly, e.g., being sensitive to the changes of prompts when extracting factual knowledge. To mitigate this issue, we propose to let PLMs learn the deterministic relationship between the remaining context and the masked content. The deterministic relationship ensures that the masked factual content can be deterministically inferable based on the existing clues in the context. That would provide more stable patterns for PLMs to capture factual knowledge than randomly masking. Two pre-training tasks are further introduced to motivate PLMs to rely on the deterministic relationship when filling masks. Specifically, we use an external Knowledge Base (KB) to identify deterministic relationships and continuously pre-train PLMs with the proposed methods. The factual knowledge probing experiments indicate that the continuously pre-trained PLMs achieve better robustness in factual knowledge capturing. Further experiments on question-answering datasets show that trying to learn a deterministic relationship with the proposed methods can also help other knowledge-intensive tasks., Accepted at EMNLP 2022

MIRACL (Multilingual Information Retrieval Across a Continuum of Languages) is a multilingual dataset we have built for the WSDM 2023 Cup challenge that focuses on ad hoc retrieval across 18 different languages, which collectively encompass over three billion native speakers around the world. These languages have diverse typologies, originate from many different language families, and are associated with varying amounts of available resources -- including what researchers typically characterize as high-resource as well as low-resource languages. Our dataset is designed to support the creation and evaluation of models for monolingual retrieval, where the queries and the corpora are in the same language. In total, we have gathered over 700k high-quality relevance judgments for around 77k queries over Wikipedia in these 18 languages, where all assessments have been performed by native speakers hired by our team. Our goal is to spur research that will improve retrieval across a continuum of languages, thus enhancing information access capabilities for diverse populations around the world, particularly those that have been traditionally underserved. This overview paper describes the dataset and baselines that we share with the community. The MIRACL website is live at http://miracl.ai/.

Collecting supporting evidence from large corpora of text (e.g., Wikipedia) is of great challenge for open-domain Question Answering (QA). Especially, for multi-hop open-domain QA, scattered evidence pieces are required to be gathered together to support the answer extraction. In this paper, we propose a new retrieval target, hop, to collect the hidden reasoning evidence from Wikipedia for complex question answering. Specifically, the hop in this paper is defined as the combination of a hyperlink and the corresponding outbound link document. The hyperlink is encoded as the mention embedding which models the structured knowledge of how the outbound link entity is mentioned in the textual context, and the corresponding outbound link document is encoded as the document embedding representing the unstructured knowledge within it. Accordingly, we build HopRetriever which retrieves hops over Wikipedia to answer complex questions. Experiments on the HotpotQA dataset demonstrate that HopRetriever outperforms previously published evidence retrieval methods by large margins. Moreover, our approach also yields quantifiable interpretations of the evidence collection process., Accepted at AAAI 2021

Zerovalent iron (ZVI) has been a major focus of research and has attracted great attention during the last 2 decades by international researchers because of its excellent pollutant removal performance and several other merits in environmental remediation. Based on Web of Science Core Collection data, we present a comprehensive bibliometric analysis of ZVI research from 2000 to 2019. We analyze 4472 publications assuming three stages of growth trend of annual publication totals. We find that “The Chemical Engineering Journal” has been the most productive journal; Noubactep C is identified as the most productive author; China has been the most active country in this field and the Chinese Academy of Science the most productive institution. The timeline of keywords shows seven distinct co-citation clusters. In addition, the top 38 keywords with strong citation bursts are also detected, suggesting that the innovation of green composite synthesis of ZVI and nanoscale ZVI and its efficient removal capacity might be the prevailing research directions in the future.

Background Cataracts are defects of the lens that cause progressive visual impairment and ultimately blindness in many vertebrate species. Most cataracts are age-related, but up to one third have an underlying genetic cause. Cataracts are common in captive zoo animals, but it is often unclear whether these are congenital or acquired (age-related) lesions. Results Here we used a functional candidate gene screening approach to identify mutations associated with cataracts in a captive giant panda (Ailuropoda melanoleuca). We screened 11 genes often associated with human cataracts and identified a novel missense mutation (c.686G > A) in the MIP gene encoding major intrinsic protein. This is expressed in the lens and normally accumulates in the plasma membrane of lens fiber cells, where it plays an important role in fluid transport and cell adhesion. The mutation causes the replacement of serine with asparagine (p.S229N) in the C-terminal tail of the protein, and modeling predicts that the mutation induces conformational changes that may interfere with lens permeability and cell–cell interactions. Conclusion The c.686G > A mutation was found in a captive giant panda with a unilateral cataract but not in 18 controls from diverse regions in China, suggesting it is most likely a genuine disease-associated mutation rather than a single-nucleotide polymorphism. The mutation could therefore serve as a new genetic marker to predict the risk of congenital cataracts in captive giant pandas.

The image deblurring technology of dynamic scene is a challenging computer vision problem. The blurry images not only affect the subjective perception but also affect the performance of the subsequent intelligent analysis. An image deblurring method of dynamic scene based on attention residual CODEC network was proposed. Firstly, in the coding stage, many residual modules were used to extract the features, and the spatial attention module was added to perceive the blurry spatial position information. Then, a global-local residual connection strategy in the internet was adopted to fuse the multi-layer convolution features to reduce the information loss. Finally, a restored image with clear edges and structure was generated in the decoding stage. The experimental results show that the peak signal-to-noise ratio (PSNR) obtained on the public data set is 31.76 dB, and the structural similarity index measure (SSIM) value is 0.912. Both the objective and subjective quality evaluations indicate that the proposed method can effectively recover the clear images containing abundant edge contour information, which obtains the optimal performance in the compared algorithm.

Modern medicine tells us that the human body is an organism composed of heart, lung, liver, kidney, spleen, stomach, brain, nerves, muscles, bones, blood vessels, blood and so on, while Traditional Chinese Medicine believes that besides these tissues and organs

Increasing amounts of information force the continuous improvement of information storage and processing technologies, further device miniaturization, and their efficiency increase. Magnetic skyrmions, topological quasiparticles, and the smallest stable magnetic textures possess intriguing properties and potential for data storage applications. Hybrid nanostructures with elements of different magnetization orientations can offer additional advantages for developing skyrmion-based spintronic and magnonic devices. We show that an N\'eel-type skyrmion confined within a nanodot placed on top of a ferromagnetic stripe produces a unique and compelling platform for exploring mutual coupling between magnetization textures. The skyrmion induces an imprint upon the stripe, which, in turn, asymmetrically squeezes the skyrmion in the dot, increasing their size and the range of skyrmion stability for small values of DMI, as well as introducing skyrmion bi-stability. At the end, we present a proof-of-concept technique for unconstrained transport of a skyrmion along a racetrack based on proposed hybrid systems. Our results demonstrate a hybrid structure that is promising for applications in magnonics and spintronics., Comment: 31 pages, 11 figures

Although achieving significant progress, existing deep generative inpainting methods are far from real-world applications due to the low generalization across different scenes. As a result, the generated images usually contain artifacts or the filled pixels differ greatly from the ground truth. Image-level predictive filtering is a widely used image restoration technique, predicting suitable kernels adaptively according to different input scenes. Inspired by this inherent advantage, we explore the possibility of addressing image inpainting as a filtering task. To this end, we first study the advantages and challenges of image-level predictive filtering for image inpainting: the method can preserve local structures and avoid artifacts but fails to fill large missing areas. Then, we propose semantic filtering by conducting filtering on the deep feature level, which fills the missing semantic information but fails to recover the details. To address the issues while adopting the respective advantages, we propose a novel filtering technique, i.e., Multilevel Interactive Siamese Filtering (MISF), which contains two branches: kernel prediction branch (KPB) and semantic & image filtering branch (SIFB). These two branches are interactively linked: SIFB provides multi-level features for KPB while KPB predicts dynamic kernels for SIFB. As a result, the final method takes the advantage of effective semantic & image-level filling for high-fidelity inpainting. We validate our method on three challenging datasets, i.e., Dunhuang, Places2, and CelebA. Our method outperforms state-of-the-art baselines on four metrics, i.e., L1, PSNR, SSIM, and LPIPS. Please try the released code and model at https://github.com/tsingqguo/misf., Accepted by CVPR 2022

Autoimmune bullous diseases (AIBDs), presenting cutaneous and/or mucosal bullous lesions, are classified into pemphigus and pemphigoid diseases. A longtime observation for complicated AIBD cases is rarely reported. In this study, serum samples of one AIBD patient were collected at seven different time points during the disease course including a relapse, which were examined by our conventional and newly developed methods for the detection of autoantibodies. Interestingly, we found changes of both the presence and the titers of various autoantibodies in accordance with the changes of clinical features during the whole disease course, which indicated that the patient started as bullous pemphigoid and relapsed as concurrence of bullous pemphigoid and mucosal-dominant-type pemphigus vulgaris.

Exhaled breath condensate (EBC) is increasingly being used for disease diagnosis and environmental exposure assessment as a noninvasive method reducing the risk of exposure. The purpose of this study was to investigate the application of a new sample type of EBC in pneumonia by metabolomics and to explore differential metabolites and potential metabolic pathways. A case-control study was performed at the Peking University Third Hospital from August to December 2020. C-MS/MS analyses were performed on EBC samples using a UHPLC system. Totally 22 patients with pneumonia and 24 healthy controls were recruited. Using untargeted metabolomics based on LC-MS/MS analysis, 25 kinds of differential metabolites were found. Through a comprehensive analysis of the pathways in which the differential metabolites were located, the key pathway with the highest correlation with the difference of metabolites was taurine and hypotaurine metabolism. The study implicates that the hypotaurine/taurine metabolic pathway may play a role on the development of pneumonia through metabolism analysis on EBC and the 3-Sulfinoalanine may be used as a biomarker in the diagnosis of pneumonia.

Supplemental material, sj-docx-1-wes-10.1177_09500170221133714 for Education-Occupation Linkage in the Highly-Educated Workforce: Patterns and Sources of Difference by Race/Ethnicity by Yao Lu, Xiaoguang Li and Benjamin Elbers in Work, Employment and Society

Supplemental material, sj-docx-1-wes-10.1177_09500170221133714 for Education-Occupation Linkage in the Highly-Educated Workforce: Patterns and Sources of Difference by Race/Ethnicity by Yao Lu, Xiaoguang Li and Benjamin Elbers in Work, Employment and Society

Increasing amounts of information force the continuous improvement of information storage and processing technologies, further device miniaturization, and their efficiency increase. Magnetic skyrmions, topological quasiparticles, and the smallest stable magnetic textures possess intriguing properties and potential for data storage applications. Hybrid nanostructures with elements of different magnetization orientations can offer additional advantages for developing skyrmion-based spintronic and magnonic devices. We show that an Néel-type skyrmion confined within a nanodot placed on top of a ferromagnetic stripe produces a unique and compelling platform for exploring mutual coupling between magnetization textures. The skyrmion induces an imprint upon the stripe, which, in turn, asymmetrically squeezes the skyrmion in the dot, increasing their size and the range of skyrmion stability for small values of DMI, as well as introducing skyrmion bi-stability. At the end, we present a proof-of-concept technique for unconstrained transport of a skyrmion along a racetrack based on proposed hybrid systems. Our results demonstrate a hybrid structure that is promising for applications in magnonics and spintronics., 31 pages, 11 figures

To alleviate the data scarcity problem in training question answering systems, recent works propose additional intermediate pre-training for dense passage retrieval (DPR). However, there still remains a large discrepancy between the provided upstream signals and the downstream question-passage relevance, which leads to less improvement. To bridge this gap, we propose the HyperLink-induced Pre-training (HLP), a method to pre-train the dense retriever with the text relevance induced by hyperlink-based topology within Web documents. We demonstrate that the hyperlink-based structures of dual-link and co-mention can provide effective relevance signals for large-scale pre-training that better facilitate downstream passage retrieval. We investigate the effectiveness of our approach across a wide range of open-domain QA datasets under zero-shot, few-shot, multi-hop, and out-of-domain scenarios. The experiments show our HLP outperforms the BM25 by up to 7 points as well as other pre-training methods by more than 10 points in terms of top-20 retrieval accuracy under the zero-shot scenario. Furthermore, HLP significantly outperforms other pre-training methods under the other scenarios., Accepted by ACL 2022 main conference; The dataset and code are available at https://github.com/jzhoubu/HLP

Although local differential privacy (LDP) protects individual users' data from inference by an untrusted data curator, recent studies show that an attacker can launch a data poisoning attack from the user side to inject carefully-crafted bogus data into the LDP protocols in order to maximally skew the final estimate by the data curator. In this work, we further advance this knowledge by proposing a new fine-grained attack, which allows the attacker to fine-tune and simultaneously manipulate mean and variance estimations that are popular analytical tasks for many real-world applications. To accomplish this goal, the attack leverages the characteristics of LDP to inject fake data into the output domain of the local LDP instance. We call our attack the output poisoning attack (OPA). We observe a security-privacy consistency where a small privacy loss enhances the security of LDP, which contradicts the known security-privacy trade-off from prior work. We further study the consistency and reveal a more holistic view of the threat landscape of data poisoning attacks on LDP. We comprehensively evaluate our attack against a baseline attack that intuitively provides false input to LDP. The experimental results show that OPA outperforms the baseline on three real-world datasets. We also propose a novel defense method that can recover the result accuracy from polluted data collection and offer insight into the secure LDP design.

The vertical ultimate bearing capacity of a single pile is an important parameter required in the design of pile foundation. The vertical compressive static load test of a single pile is the most reliable method to determine the vertical ultimate bearing capacity of a single pile. However, the load test cost is high and the cycle is long. In the early stage of small and medium-sized projects or large-scale projects, the conditions for carrying out the vertical compressive static load test of a single pile are generally not available. Based on the geotechnical engineering of a coal-fired power station in Southeast Asia, the vertical ultimate bearing capacity of a single pile is estimated by using the results of cone penetration test (CPT) and standard penetration test (SPT), and the ultimate bearing capacity is tested by high strain dynamic test. Compare various estimating value and high strain dynamic testing value to static load test value with ultimate vertical bearing capacity of a single pile, and then obtain the reliability of vertical ultimate bearing capacity of a single pile by means of CPT and SPT in this area. By means of the research results of this paper, obtain the vertical ultimate bearing capacity of PHC pipe pile quickly and accurately in this area for similar engineering design. Ultimately achieving the goal of shortening the construction period and reducing geotechnical investigation cost.

The monitoring of harmful phytoplankton is very important for the maintenance of the aquatic ecological environment. Traditional algae monitoring methods require professionals with substantial experience in algae species, which are time-consuming, expensive, and limited in practice. The automatic classification of algae cell images and the identification of harmful phytoplankton images were realized by the combination of multiple convolutional neural networks (CNNs) and deep learning techniques based on transfer learning in this work. Eleven common harmful and 31 harmless phytoplankton genera were collected as input samples; the five CNNs classification models of AlexNet, VGG16, GoogLeNet, ResNet50, and MobileNetV2 were fine-tuned to automatically classify phytoplankton images; and the average accuracy was improved 11.9% when compared to models without fine-tuning. In order to monitor harmful phytoplankton which can cause red tides or produce toxins severely polluting drinking water, a new identification method of harmful phytoplankton which combines the recognition results of five CNN models was proposed, and the recall rate reached 98.0%. The experimental results validate that the recognition performance of harmful phytoplankton could be significantly improved by transfer learning, and the proposed identification method is effective in the preliminary screening of harmful phytoplankton and greatly reduces the workload of professional personnel.

Bioelectrochemical systems (BESs) have great potential for treating wastewater containing polycyclic aromatic hydrocarbons (PAHs); however, detailed data on cell physiological activities in PAH biodegradation pathways stimulated by BESs are still lacking. In this paper, a novel BES device was assembled to promote the growth of Pseudomonas sp. DGYH-12 in phenanthrene (PHE) degradation. The results showed that in the micro-electric field (0.2 V), cell growth rate and PHE degradation efficiency were 22% and 27.2% higher than biological control without electric stimulation (BC), respectively. The extracellular polymeric substance (EPS) concentration in BES (39.38 mg L−1) was higher than control (33.36 mg L−1); moreover, the membrane permeability and ATPase activities were also enhanced and there existing phthalic acid and salicylic acid metabolic pathways in the strain. The degradation genes nahAc, pcaH, and xylE expression levels were upregulated by micro-electric stimulation. This is the first study to analyze the physiological and metabolic effect of micro-electric stimulation on a PHE-degrading strain in detail and systematically. Electronic supplementary material The online version of this article (10.1007/s11356-019-05670-5) contains supplementary material, which is available to authorized users.

High-performance Zn-Al layered double hydroxide nanosheets-coated spherical ZnO@C composite is successfully fabricated for use as advanced anode materials of zinc-based rechargeable batteries. It is interestingly found that the spherical ZnO@C-ZnAl layered double hydroxide (15%) composite with a proper coated content of ZnAl layered double hydroxide shows a superior cycling performance, satisfactory volume specific capacity and enhanced high-rate properties. After 400 cycles at 1C/1C and 200 cycles at 1C/5C, the ZnO@C-ZnAl layered double hydroxide (15%) exhibits enhanced capacity retention of 81.6%. In contrast, the pristine ZnO shows low capacity retention of only 46.3% after 200 cycles at 1C/1C. These performance improvements can be mainly attributable to the synergistic effect of the surface coating with Zn-Al layered double hydroxide and in-situ doping with carbon. The coated Zn-Al layered double hydroxide nanosheets can modify the“dissolution-deposition” process of ZnO anode and the doped C can enhance its electronic conductivity, eventually leading to the reduced electrochemical resistance, improved reaction reversibility, and enhanced structural stability. As a result, it is believed that the spherical ZnO@C-ZnAl layered double hydroxide (15%) composite is a promising advanced anode material for zinc-based rechargeable batteries.

Bone marrow mesenchymal stem cells undergo differentiation to different lineages with different efficiencies when induced by different factors. We added a bFGF-chitosan controlled release system (bFGF-CCRS) as an inducer into conditioned medium to facilitate the oriented differentiation of BMSCs into neural lineage cells (eventually mature neurons); furthermore, we synchronized BMSCs to the G0/G1 phase via serum starvation to observe the effect of the inducer on the differentiation direction and efficiency. The nonsynchronized group, chitosan alone (not loaded with bFGF) group, soluble bFGF group, and conditioned medium group served as controls, and we observed the dynamic process of differentiation of BMSCs into neural lineage cells at different time points after the beginning of coculture. We analyzed the binding patterns of bFGF and chitosan and assayed the expression differences of key factors (FGFR1, ERK, and c-fos) and molecular switches (BTG2) that regulate the transformation from cell proliferation to differentiation. We also investigated the potential molecular mechanism of BMSC differentiation into neural lineage cells at a high percentage when induced by bFGF-CCRS.

Nickel nanoparticles uniformly anchored on reduced graphene oxide (Ni/RGO) nanosheets is successfully prepared by a facile wet-chemical method. It is found that many small Ni nanoparticles with the particle size of 4–8 nm are anchored on RGO, showing uniform dispersion. The Ni/RGO nanocomposite-modified glassy carbon electrode (GCE) (i.e., Ni/RGO/GCE) is investigated as an electrocatalyst for the ethanol oxidation reaction. Especially, the as-prepared Ni/RGO/GCE delivers excellent electrocatalytic performance in terms of the low peak potential value, the low reduction of its overvoltage, and satisfactory peak current density. The excellent electrocatalytic properties could be ascribed to the synergistic effect of Ni spherical nanoparticles and RGO nanosheets, which can lead to the high use ratio of Ni. It is proposed that the anodic peak that emerged in the potential negative-going sweep may be attributed to the oxidation of ethanol, which results from the high concentration of ethanol and an autocatalytic process at Ni/RGO/GCE.

Radioactive jaw necrosis is a serious complication of head and neck radiotherapy. This complication 􀆳s main characteristics are chronic necrosis and infection, and its clinical manifestations are local pain, mouth opening re⁃ striction, speech disorder, dysphagia, exposure of dead bone, prolonged healing of facial and neck soft tissue fistula, and even pathological fracture, which brings great pain to the patient􀆳s body and mind. Radioactive jaw necrosis has become a clinical problem in the field of medicine, specifically for maxillofacial surgery of the external head and neck. The pathogenesis of radiation⁃induced jaw necrosis is still unclear, but high radiotherapy dose, extraction of teeth after radio⁃ therapy and surgical trauma are recognized as high⁃risk factors. The diagnosis of radiation⁃induced jaw necrosis needs to be combined with the patient􀆳s radiotherapy history, clinical symptoms and imaging examination. Conservative or sur⁃ gical treatment should be selected according to its clinical classification and stages, and the prevention of radiation⁃in⁃ duced jaw necrosis should be emphasized. The key points are to remove dental caries, residual roots and crowns before radiotherapy; improve radiotherapy technology; and take preventive measures against high⁃risk factors. In recent years, the diagnosis and treatment of radiation⁃induced jaw necrosis has gradually become standardized and individualized. This article reviews the progress of treatment and research on radiation⁃induced jaw necrosis at home and abroad in terms of its definition, etiology, risk factors, diagnosis, treatment and prevention to provide reference for clinical treat⁃ ment.

Open-domain Question Answering (ODQA) has achieved significant results in terms of supervised learning manner. However, data annotation cannot also be irresistible for its huge demand in an open domain. Though unsupervised QA or unsupervised Machine Reading Comprehension (MRC) has been tried more or less, unsupervised ODQA has not been touched according to our best knowledge. This paper thus pioneers the work of unsupervised ODQA by formally introducing the task and proposing a series of key data construction methods. Our exploration in this work inspiringly shows unsupervised ODQA can reach up to 86% performance of supervised ones.

Modern medicine tells us that the human body is an organism composed of heart, lungs, liver, kidneys, spleen, stomach, brain, nerves, muscles, bones, blood vessels, blood and so on, but Traditional Chinese Medicine believes that besides these tissues and organs, the human body still has another part of the structure, Traditional Chinese Medicine calls them Jing Luo and Shu Xue. Jing Luo means the longitudinal line of the human body and the accompanying net, translated into English Meridians and Collaterals. Shu Xue means holes distributed on Jing Luo and outside Jing Luo, because stimulating Shu Xue's position by acupuncture, massage and other methods can cure diseases, so Shu Xue is translated into an English acupuncture point, abbreviated as acupoint or point. Meridians and acupoints are the special knowledge of the human body structure in Traditional Chinese Medicine. Traditional Chinese Medicine not only draws the distribution map of the meridians and acupoints in the human body, but also has been using them to treat diseases for thousands of years. There are hundreds of these acupoints, stimulating each one by acupuncture, massage or other methods will have a special effect on the human body and can treat various diseases. But what effect does stimulating every acupoint have on the human body so that it can treat various diseases? The discussion of Traditional Chinese Medicine is vague and incomprehensible, and can not be proved by experiments. According to the author's research for more than 30 years, this paper makes a clear and accurate exposition of the effects on the human body and diseases that can be treated with acupoint massage. These statements can be proved by experiments, so they are believed to be reliable. It is hoped that meridians, acupoints and massage therapy can be incorporated into modern medicine and become a part of modern medicine in the future after the contents of this paper be proved by others through experiments. Massaging acupoints can not only treat many diseases that are difficult to be treated with drugs, but also have simple methods and low cost.

The application of common mode and differential mode separators for electromagnetic noise is of great significance for improving the efficiency of electromagnetic compatibility trouble-shooting. By analyzing the characteristics of different electromagnetic noise separators, a common mode and differential mode separator network model for electromagnetic noise based on autotransformers is proposed. Through parameter extraction and simulation analysis, a common mode and differential mode separator for electromagnetic noise is designed. After comparing the actual measurement and simulation results of the separator parameters, the effectiveness of the separator design is verified.

Vibration energy harvesters (VEHs) can transform ambient vibration energy to electricity and have been widely investigated as promising self-powered devices for wireless sensor networks, wearable sensors, and applications of a micro-electro-mechanical system (MEMS). However, the ambient vibration is always too weak to hinder the high energy conversion efficiency. In this paper, the integrated frame composed of piezoelectric beams and mechanical amplifiers is proposed to improve the energy conversion efficiency of a VEH. First, the initial structures of a piezoelectric frame (PF) and an amplification frame (AF) are designed. The dynamic model is then established to analyze the influence of key structural parameters on the mechanical amplification factor. Finite element simulation is conducted to study the energy harvesting performance, where the stiffness characteristics and power output in the cases of series and parallel load resistance are discussed in detail. Furthermore, piezoelectric beams with variable cross-sections are introduced to optimize and improve the energy harvesting efficiency. Advantages of the PF with the AF are illustrated by comparison with conventional piezoelectric cantilever beams. The results show that the proposed integrated VEH has a good mechanical amplification capability and is more suitable for low-frequency vibration conditions.

Aiming at the problem that it is difficult to obtain the complete and timely device status information of gas-insulated switchgear for traditional methods in deformation monitoring, leading to unable to accurately control the device, which will influence the substation’s safe operation for sure, the paper proposed to use BDS independently developed by China to realize the intelligent detection of the gas-insulated switchgear, and designed a BDS deformation monitoring system for the gas-insulated switchgear of the EHV substation. The experimental result showed that the proposed system could efficiently monitor and pre-warn the deformation of the device.

Cataracts are a common cause of visual impairment and blindness in mammals. They are usually associated with aging, but approximately one third of cases have a significant genetic component. Cataracts are increasingly prevalent among aging populations of captive giant pandas (Ailuropoda melanoleuca) and it is therefore important to identify genetic determinants that influence the likelihood of cataract development in order to distinguish between congenital and age-related disease. Here we screened for cataract-related genetic effects using a functional candidate gene approach combined with bioinformatics to identify the underlying genetic defect in a giant panda with congenital cataracts. We identified a missense mutation in exon 10 of the HSF4 gene encoding heat shock transcription factor 4. The mutation causes the amino acid substitution R377W in a highly conserved segment of the protein between the isoform-specific and downstream hydrophobic regions. Predictive modeling revealed that the substitution is likely to increase the hydrophobicity of the protein and disrupt interactions with spatially adjacent amino acid side chains. The mutation was not found in 13 unaffected unrelated animals but was found in an unrelated animal also diagnosed with senile congenital cataract. The novel missense mutation in the HSF4 gene therefore provides a potential new genetic determinant that could help to predict the risk of cataracts in giant pandas.

Sentiment analysis has been widely used in text mining of social media to discover valuable information from user reviews. Sentiment lexicon is an essential tool for sentiment analysis. Recent research studies indicate that constructing sentiment lexicons for special domains can achieve better results in sentiment analysis. However, it is not easy to construct a sentiment lexicon for a specific domain because most current methods highly depend on general sentiment lexicons and complex linguistic rules. In this paper, the construction of sentiment lexicon is transformed into a training-optimization process. In our scheme, the accuracy of sentiment classification is used as the optimization objective. The candidate sentiment lexicons are regarded as the individuals that need to be optimized. Then, two genetic algorithms are specially designed to adjust the values of sentiment words in lexicon. Finally, the best individual evolved in the presented genetic algorithms is selected as the sentiment lexicon. Our method only depends on some labelled texts and does not need any linguistic knowledge or prior knowledge. It provides a simple and easy way to construct a sentiment lexicon in a specific domain. Experiment results show that the proposed method has good flexibility and can generate high-quality sentiment lexicon in specific domains.

Additional file 4.

Additional file 5.

Sequential recommender systems aim to model users' evolving interests from their historical behaviors, and hence make customized time-relevant recommendations. Compared with traditional models, deep learning approaches such as CNN and RNN have achieved remarkable advancements in recommendation tasks. Recently, the BERT framework also emerges as a promising method, benefited from its self-attention mechanism in processing sequential data. However, one limitation of the original BERT framework is that it only considers one input source of the natural language tokens. It is still an open question to leverage various types of information under the BERT framework. Nonetheless, it is intuitively appealing to utilize other side information, such as item category or tag, for more comprehensive depictions and better recommendations. In our pilot experiments, we found naive approaches, which directly fuse types of side information into the item embeddings, usually bring very little or even negative effects. Therefore, in this paper, we propose the NOninVasive self-attention mechanism (NOVA) to leverage side information effectively under the BERT framework. NOVA makes use of side information to generate better attention distribution, rather than directly altering the item embedding, which may cause information overwhelming. We validate the NOVA-BERT model on both public and commercial datasets, and our method can stably outperform the state-of-the-art models with negligible computational overheads., Comment: Accepted at AAAI 2021

Additional file 2.

Additional file 2. Differentially expressed genes with significant prognostic capacity in HNSCC.

Additional file 3.

Additional file 1: Table S1 Demographic information on included studies. Table S2 Risk factors of patients undergoing mitral valve versus replacement. Table S3. Quality ratings for the cohort studies included on the basis of Newcastle-Ottawa quality assessment scale.

Additional file 1. Overlapping genes among three public datasets.