Explore the vast range of titles available.

Background and aimsLiver fibrosis is a wound-healing response that disrupts the liver architecture and function by replacing functional parenchyma with scar tissue. Recent progress has advanced our knowledge of this scarring process, but the detailed mechanism of liver fibrosis is far from clear.MethodsThe fibrotic specimens of patients and HLF (hepatic leukemia factor)PB/PB mice were used to assess the expression and role of HLF in liver fibrosis. Primary murine hepatic stellate cells (HSCs) and human HSC line Lx2 were used to investigate the impact of HLF on HSC activation and the underlying mechanism.ResultsExpression of HLF was detected in fibrotic livers of patients, but it was absent in the livers of healthy individuals. Intriguingly, HLF expression was confined to activated HSCs rather than other cell types in the liver. The loss of HLF impaired primary HSC activation and attenuated liver fibrosis in HLFPB/PB mice. Consistently, ectopic HLF expression significantly facilitated the activation of human HSCs. Mechanistic studies revealed that upregulated HLF transcriptionally enhanced interleukin 6 (IL-6) expression and intensified signal transducer and activator of transcription 3 (STAT3) phosphorylation, thus promoting HSC activation. Coincidentally, IL-6/STAT3 signalling in turn activated HLF expression in HSCs, thus completing a feedforward regulatory circuit in HSC activation. Moreover, correlation between HLF expression and alpha-smooth muscle actin, IL-6 and p-STAT3 levels was observed in patient fibrotic livers, supporting the role of HLF/IL-6/STAT3 cascade in liver fibrosis.ConclusionsIn aggregate, we delineate a paradigm of HLF/IL-6/STAT3 regulatory circuit in liver fibrosis and propose that HLF is a novel biomarker for activated HSCs and a potential target for antifibrotic therapy.

This is the first randomised controlled trial for assessment of the immunogenicity and safety of a candidate non-replicating adenovirus type-5 (Ad5)-vectored COVID-19 vaccine, aiming to determine an appropriate dose of the candidate vaccine for an efficacy study. This randomised, double-blind, placebo-controlled, phase 2 trial of the Ad5-vectored COVID-19 vaccine was done in a single centre in Wuhan, China. Healthy adults aged 18 years or older, who were HIV-negative and previous severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection-free, were eligible to participate and were randomly assigned to receive the vaccine at a dose of 1 × 1011 viral particles per mL or 5 × 1010 viral particles per mL, or placebo. Investigators allocated participants at a ratio of 2:1:1 to receive a single injection intramuscularly in the arm. The randomisation list (block size 4) was generated by an independent statistician. Participants, investigators, and staff undertaking laboratory analyses were masked to group allocation. The primary endpoints for immunogenicity were the geometric mean titres (GMTs) of specific ELISA antibody responses to the receptor binding domain (RBD) and neutralising antibody responses at day 28. The primary endpoint for safety evaluation was the incidence of adverse reactions within 14 days. All recruited participants who received at least one dose were included in the primary and safety analyses. This study is registered with ClinicalTrials.gov, NCT04341389. 603 volunteers were recruited and screened for eligibility between April 11 and 16, 2020. 508 eligible participants (50% male; mean age 39·7 years, SD 12·5) consented to participate in the trial and were randomly assigned to receive the vaccine (1 × 1011 viral particles n=253; 5 × 1010 viral particles n=129) or placebo (n=126). In the 1 × 1011 and 5 × 1010 viral particles dose groups, the RBD-specific ELISA antibodies peaked at 656·5 (95% CI 575·2–749·2) and 571·0 (467·6–697·3), with seroconversion rates at 96% (95% CI 93–98) and 97% (92–99), respectively, at day 28. Both doses of the vaccine induced significant neutralising antibody responses to live SARS-CoV-2, with GMTs of 19·5 (95% CI 16·8–22·7) and 18·3 (14·4–23·3) in participants receiving 1 × 1011 and 5 × 1010 viral particles, respectively. Specific interferon γ enzyme-linked immunospot assay responses post vaccination were observed in 227 (90%, 95% CI 85–93) of 253 and 113 (88%, 81–92) of 129 participants in the 1 × 1011 and 5 × 1010 viral particles dose groups, respectively. Solicited adverse reactions were reported by 183 (72%) of 253 and 96 (74%) of 129 participants in the 1 × 1011 and 5 × 1010 viral particles dose groups, respectively. Severe adverse reactions were reported by 24 (9%) participants in the 1 × 1011 viral particles dose group and one (1%) participant in the 5 × 1010 viral particles dose group. No serious adverse reactions were documented. The Ad5-vectored COVID-19 vaccine at 5 × 1010 viral particles is safe, and induced significant immune responses in the majority of recipients after a single immunisation. National Key R&D Programme of China, National Science and Technology Major Project, and CanSino Biologics.

Hyperspectral images (HSIs) data that is typically presented in 3-D format offers an opportunity for 3-D networks to extract spectral and spatial features simultaneously. In this paper, we propose a novel end-to-end 3-D dense convolutional network with spectral-wise attention mechanism (MSDN-SA) for HSI classification. The proposed MSDN-SA exploits 3-D dilated convolutions to simultaneously capture the spectral and spatial features at different scales, and densely connects all 3-D feature maps with each other. In addition, a spectral-wise attention mechanism is introduced to enhance the distinguishability of spectral features, which improves the classification performance of the trained models. Experimental results on three HSI datasets demonstrate that our MSDN-SA achieves competitive performance for HSI classification.

Upper protective seam mining has been widely applied in China, but the theory of long-distance multiple upper protective seam mining is not yet perfect. In order to investigate the overburden stress evolution law of repetitive mining of long-distance coal seam groups, an experimental study was conducted to simulate similar materials under repeated mining conditions in the long-distance double upper protective layer in the background of Pingmei Group 8th coal mine. By analyzing the roof-collapse structure and the stress evolution law of different layers of the floor during the superposition mining, the pressure-relief range of the protective layer after the mining of the double upper protective layer was determined. The study results showed that: the pressure relief of the protective layer in the long-distance upper protective layer mining was a dynamic process. After the mining of Group D coal seam, the maximum impact depth of the bottom plate could reach 182 m, and the pressure-relief angle of the upper side of Group E coal seam was 65°, and the pressure-relief angle of the lower side was 75°. The distance behind the vertical projection of the working face of Group D was 42 m. The overlapping back mining would affect the stress distribution of Group F coal seam. The pressure-relief angle of the upper side of Group F coal seam was 88°, and the pressure-relief angle of the lower side was greater than 78°. The distance behind the vertical projection of the working face of Group E was less than 61 m. The superposition and staggered mining of double protective layers could expand the protective layer. Through the verification of the measurement of gas parameters on site, it can be seen from the results that it has a certain protection effect. The research results can enrich the theory of long-distance multiple upper protective layer mining, and provide theoretical guidance for long-distance Coal Seam Group Mining in Pingmei coal-mine area.

In recent years, convolutional neural networks (CNNs) have played a dominant role in facial expression recognition. While CNN-based methods have achieved remarkable success, they are notorious for having an excessive number of parameters, and they rely on a large amount of manually annotated data. To address this challenge, we expand the number of training samples by learning expressions from a face recognition dataset to reduce the impact of a small number of samples on the network training. In the proposed deep joint learning framework, the deep features of the face recognition dataset are clustered, and simultaneously, the parameters of an efficient CNN are learned, thereby marking the data for network training automatically and efficiently. Specifically, first, we develop a new efficient CNN based on the proposed affinity convolution module with much lower computational overhead for deep feature learning and expression classification. Then, we develop an expression-guided deep facial clustering approach to cluster the deep features and generate abundant expression labels from the face recognition dataset. Finally, the AC-based CNN is fine-tuned using an updated training set and a combined loss function. Our framework is evaluated on several challenging facial expression recognition datasets as well as a self-collected dataset. In the context of facial expression recognition applied to the field of education, our proposed method achieved an impressive accuracy of 95.87% on the self-collected dataset, surpassing other existing methods.

Hyperspectral image (HSI) classification has been marked by exceptional progress in recent years. Much of this progess has come from advances in convolutional neural networks (CNNs). Different from the RGB images, HSI images are captured by various remote sensors with different spectral configurations. Moreover, each HSI dataset only contains very limited training samples and thus the model is prone to overfitting when using deep CNNs. In this paper, we first propose a 3D asymmetric inception network, AINet, to overcome the overfitting problem. With the emphasis on spectral signatures over spatial contexts of HSI data, the 3D convolution layer of AINet is replaced with two asymmetric inception units, i.e., a space inception unit and spectrum inception unit, to convey and classify the features effectively. In addition, we exploited a data-fusion transfer learning strategy to improve model initialization and classification performance. Extensive experiments show that the proposed approach outperforms all of the state-of-the-art methods via several HSI benchmarks, including Pavia University, Indian Pines and Kennedy Space Center (KSC).

The recognition of human activity is crucial as the Internet of Things (IoT) progresses toward future smart homes. Wi-Fi-based motion-recognition stands out due to its non-contact nature and widespread applicability. However, the channel state information (CSI) related to human movement in indoor environments changes with the direction of movement, which poses challenges for existing Wi-Fi movement-recognition methods. These challenges include limited directions of movement that can be detected, short detection distances, and inaccurate feature extraction, all of which significantly constrain the wide-scale application of Wi-Fi action-recognition. To address this issue, we propose a direction-independent CSI fusion and sharing model named CSI-F, one which combines Convolutional Neural Networks (CNN) and Gated Recurrent Units (GRU). Specifically, we have introduced a series of signal-processing techniques that utilize antenna diversity to eliminate random phase shifts, thereby removing noise influences unrelated to motion information. Later, by amplifying the Doppler frequency shift effect through cyclic actions and generating a spectrogram, we further enhance the impact of actions on CSI. To demonstrate the effectiveness of this method, we conducted experiments on datasets collected in natural environments. We confirmed that the superposition of periodic actions on CSI can improve the accuracy of the process. CSI-F can achieve higher recognition accuracy compared with other methods and a monitoring coverage of up to 6 m.

Recently, Hyperspectral Image (HSI) classification methods based on deep learning models have shown encouraging performance. However, the limited numbers of training samples, as well as the mixed pixels due to low spatial resolution, have become major obstacles for HSI classification. To tackle these problems, we propose a resource-efficient HSI classification framework which introduces adaptive spectral unmixing into a 3D/2D dense network with early-exiting strategy. More specifically, on one hand, our framework uses a cascade of intermediate classifiers throughout the 3D/2D dense network that is trained end-to-end. The proposed 3D/2D dense network that integrates 3D convolutions with 2D convolutions is more capable of handling spectral-spatial features, while containing fewer parameters compared with the conventional 3D convolutions, and further boosts the network performance with limited training samples. On another hand, considering the existence of mixed pixels in HSI data, the pixels in HSI classification are divided into hard samples and easy samples. With the early-exiting strategy in these intermediate classifiers, the average accuracy can be improved by reducing the amount of computation cost for easy samples, thus focusing on classifying hard samples. Furthermore, for hard samples, an adaptive spectral unmixing method is proposed as a complementary source of information for classification, which brings considerable benefits to the final performance. Experimental results on four HSI benchmark datasets demonstrate that the proposed method can achieve better performance than state-of-the-art deep learning-based methods and other traditional HSI classification methods.

This paper studies the classification problem of hyperspectral image (HSI). Inspired by the great success of deep neural networks in Artificial Intelligence (AI), researchers have proposed different deep learning based algorithms to improve the performance of hyperspectral classification. However, deep learning based algorithms always require a large-scale annotated dataset to provide sufficient training. To address this problem, we propose a semi-supervised deep learning framework based on the residual networks (ResNets), which use very limited labeled data supplemented by abundant unlabeled data. The core of our framework is a novel dual-strategy sample selection co-training algorithm, which can successfully guide ResNets to learn from the unlabeled data by making full use of the complementary cues of the spectral and spatial features in HSI classification. Experiments on the benchmark HSI dataset and real HSI dataset demonstrate that, with a small number of training data, our approach achieves competitive performance with maximum improvement of 41% (compare with traditional convolutional neural network (CNN) with 5 initial training samples per class on Indian Pines dataset) for HSI classification as compared with the results from those state-of-the-art supervised and semi-supervised methods.

If and how proteasomes catalyze not only peptide hydrolysis but also peptide splicing is an open question that has divided the scientific community. The debate has so far been based on immunopeptidomics, in vitro digestions of synthetic polypeptides as well as ex vivo and in vivo experiments, which could only indirectly describe proteasome-catalyzed peptide splicing of full-length proteins. Here we develop a workflow—and cognate software - to analyze proteasome-generated non-spliced and spliced peptides produced from entire proteins and apply it to in vitro digestions of 15 proteins, including well-known intrinsically disordered proteins such as human tau and α-Synuclein. The results confirm that 20S proteasomes produce a sizeable variety of cis -spliced peptides, whereas trans -spliced peptides are a minority. Both peptide hydrolysis and splicing produce peptides with well-defined characteristics, which hint toward an intricate regulation of both catalytic activities. At protein level, both non-spliced and spliced peptides are not randomly localized within protein sequences, but rather concentrated in hotspots of peptide products, in part driven by protein sequence motifs and proteasomal preferences. At sequence level, the different peptide sequence preference of peptide hydrolysis and peptide splicing suggests a competition between the two catalytic activities of 20S proteasomes during protein degradation. Do proteasomes catalyze peptide splicing? Here, the authors develop and apply a method to identify spliced peptides produced from entire proteins, confirm that proteasomes produce a sizeable variety of cis -spliced peptides with well-defined characteristics, and show that non-spliced and spliced peptides are concentrated in hotspots.