Search

Sequence recommendation is a prominent research area within recommender systems, focused on predicting items that users may be interested in by modeling their historical interaction sequences. However, due to data sparsity, user interaction sequences in sequence recommendation are typically short. A common approach to address this issue is filling sequences with zero values, significantly reducing the effective utilization of input space. Furthermore, traditional sequence recommendation methods based on self-attention mechanisms exhibit quadratic complexity with respect to sequence length. These issues affect the performance of recommendation algorithms. To tackle these challenges, we propose a multi-task sequence recommendation model, Blin, which integrates bidirectional KL divergence and linear attention. Blin abandons the conventional zero-padding strategy, opting instead for random repeat padding to enhance sequence data. Additionally, bidirectional KL divergence loss is introduced as an auxiliary task to regularize the probability distributions obtained from different sequence representations. To improve the computational efficiency compared to traditional attention mechanisms, a linear attention mechanism is employed during sequence encoding, significantly reducing the computational complexity while preserving the learning capacity of traditional attention. Experimental results on multiple public datasets demonstrate the effectiveness of the proposed model. [ABSTRACT FROM AUTHOR]

Mutations in the C-terminal of KIF1A (Kinesin family member 1A) may lead to amyotrophic lateral sclerosis (ALS) through unknown mechanisms that are not yet understood. Using iPSC reprogramming technology and motor neuron differentiation techniques, we generated iPSCs from a healthy donor and two ALS patients with KIF1A mutations (R1457Q and P1688L) and differentiated them into spinal motor neurons (iPSC-MN) to investigate KIF1A-related ALS pathology. Our in vitro iPSC-iMN model faithfully recapitulated specific aspects of the disease, such as neurite fragmentation. Through this model, we observed that these mutations led to KIF1A aggregation at the proximal axon of motor neurons and abnormal accumulation of its transport cargo, LAMP1, resulting in autophagy dysfunction and cell death. RNAseq analysis also indicated that the functions of the extracellular matrix, structure, and cell adhesion were significantly disturbed. Notably, using rapamycin during motor neuron differentiation can effectively prevent motor neuron death. [ABSTRACT FROM AUTHOR]

Simple Summary: This study investigates the potential of small extracellular vesicles (sEVs) as vehicles for delivering therapeutic agents, with a focus on the p53 tumor suppressor gene. Fusion with CD63 or ARRDC1 was found to facilitate the overexpression and proper localization of p53 and significantly increase the yield and loading efficiency of sEVs with p53 mRNA and proteins. Moreover, the functional assessment of these engineered sEVs on H1299 cells demonstrated an enhanced anti-tumor effect, particularly with ARP-sEVs, suggesting a superior inhibition of cell proliferation and promotion of apoptosis. These results are expected to be informative in the application of EV therapy. Small extracellular vesicles (sEVs) have emerged as promising therapeutic agents and drug delivery vehicles. Targeted modification of sEVs and their contents using genetic modification strategies is one of the most popular methods. This study investigated the effects of p53 fusion with arrestin domain-containing protein 1 (ARRDC1) and CD63 on the generation of sEVs, p53 loading efficiency, and therapeutic efficacy. Overexpression of either ARRDC1–p53 (ARP) or CD63–p53 (CDP) significantly elevated p53 mRNA and protein levels. The incorporation of ARRDC1 and CD63 significantly enhanced HEK293T-sEV biogenesis, evidenced by significant increases in sEV-associated proteins TSG101 and LAMP1, resulting in a boost in sEV production. Importantly, fusion with ARRDC1 or CD63 substantially increased the efficiency of loading both p53 fusion proteins and its mRNA into sEVs. sEVs equipped with ARP or CDP significantly enhanced the enrichment of p53 fusion proteins and mRNA in p53-null H1299 cells, resulting in a marked increase in apoptosis and a reduction in cell proliferation, with ARP-sEVs demonstrating greater effectiveness than CDP-sEVs. These findings underscore the enhanced functionality of ARRDC1- and CD63-modified sEVs, emphasizing the potential of genetic modifications in sEV-based therapies for targeted cancer treatment. [ABSTRACT FROM AUTHOR]

Background: After the adjustment of COVID-19 epidemic policy, mainland China experienced two consecutive waves of Omicron variants within a seven-month period. In Guangzhou city, as one of the most populous regions, the viral infection characteristics, molecular epidemiology, and the dynamic of population immunity are still elusive. Methods: We launched a prospective cohort study in the Guangdong Provincial CDC from December 2022 to July 2023. Fifty participants who received the same vaccination regimen and had no previous infection were recruited. Results: 90% of individuals were infected with Omicron BA.5* variants within three weeks in the first wave. Thirteen cases (28.26%) experienced infection with XBB.1* variants, occurring from 14 weeks to 21 weeks after the first wave. BA.5* infections exhibited higher viral loads in nasopharyngeal sites compared to oropharyngeal sites. Compared to BA.5* infections, the XBB.1* infections had significantly milder clinical symptoms, lower viral loads, and shorter durations of virus positivity. The infection with the BA.5* variant elicited varying levels of neutralizing antibodies against XBB.1* among different individuals, even with similar levels of BA.5* antibodies. The level of neutralizing antibodies specific to XBB.1* determined the risk of reinfection. Conclusions: The rapid large-scale infections of the Omicron variants have quickly established herd immunity among the population in mainland China. In the future of the COVID-19 epidemic, a lower infection rate but a longer duration can be expected. Given the large population size and ongoing diversified herd immunity, it remains crucial to closely monitor the molecular epidemiology of SARS-CoV-2 for the emergence of new variants of concern in this region. Additionally, the timely evaluation of the immune status across different age groups is essential for informing future vaccination strategies and intervention policies. [ABSTRACT FROM AUTHOR]

The electroencephalogram (EEG) has recently emerged as a pivotal tool in brain imaging analysis, playing a crucial role in accurately interpreting brain functions and states. To address the problem that the presence of ocular artifacts in the EEG signals of patients with obstructive sleep apnea syndrome (OSAS) severely affects the accuracy of sleep staging recognition, we propose a method that integrates a support vector machine (SVM) with genetic algorithm (GA)-optimized variational mode decomposition (VMD) and second-order blind identification (SOBI) for the removal of ocular artifacts from single-channel EEG signals. The SVM is utilized to identify artifact-contaminated segments within preprocessed single-channel EEG signals. Subsequently, these signals are decomposed into variational modal components across different frequency bands using the GA-optimized VMD algorithm. These components undergo further decomposition via the SOBI algorithm, followed by the computation of their approximate entropy. An approximate entropy threshold is set to identify and remove components laden with ocular artifacts. Finally, the signal is reconstructed using the inverse SOBI and VMD algorithms. To validate the efficacy of our proposed method, we conducted experiments utilizing both simulated data and real OSAS sleep EEG data. The experimental results demonstrate that our algorithm not only effectively mitigates the presence of ocular artifacts but also minimizes EEG signal distortion, thereby enhancing the precision of sleep staging recognition based on the EEG signals of OSAS patients. [ABSTRACT FROM AUTHOR]

Accurate prediction of solar irradiance holds significant value for renewable energy usage and power grid management. However, traditional forecasting methods often overlook the time dependence of solar irradiance sequences and the varying importance of different influencing factors. To address this issue, this study proposes a dual-path information fusion and twin attention-driven solar irradiance forecasting model. The proposed framework comprises three components: a residual attention temporal convolution block (RACB), a dual-path information fusion module (DIFM), and a twin self-attention module (TSAM). These components collectively enhance the performance of multi-step solar irradiance forecasting. First, the RACB is designed to enable the network to adaptively learn important features while suppressing irrelevant ones. Second, the DIFM is implemented to reinforce the model's robustness against input data variations and integrate multi-scale features. Lastly, the TSAM is introduced to extract long-term temporal dependencies from the sequence and facilitate multi-step prediction. In the solar irradiance forecasting experiments, the proposed model is compared with six benchmark models across four datasets. In the one-step predictions, the average performance metrics RMSE, MAE, and MAPE of the four datasets decreased within the ranges of 0.463–2.390 W / m 2 , 0.439–2.005 W / m 2 , and 1.3–9.2%, respectively. Additionally, the average R 2 value across the four datasets increased by 0.008 to 0.059. The experimental results indicate that the model proposed in this study exhibits enhanced accuracy and robustness in predictive performance, making it a reliable alternative for solar irradiance forecasting. [ABSTRACT FROM AUTHOR]

Fault alarm time lag is one of the difficulties in fault diagnosis of wind turbine generators (WTGs), and the existing methods are insufficient to achieve accurate and rapid fault diagnosis of WTGs, and the operation and maintenance costs of WTGs are too high. To invent a new method for fast and accurate fault diagnosis of WTGs, this study constructs a stacking integration model based on the machine learning algorithms light gradient boosting machine (LightGBM), extreme gradient boosting (XGBoost), and stochastic gradient descent regressor (SGDRegressor) using publicly available datasets from Energias De Portugal (EDP). This model is automatically tuned for hyperparameters during training using Bayesian tuning, and the coefficient of determination (R2) and root mean square error (RMSE) were used to evaluate the model to determine its applicability and accuracy. The fitted residuals of the test set were calculated, the Pauta criterion (3σ) and the temporal sliding window were applied, and a final adaptive threshold method for accurate fault diagnosis and alarming was created. The model validation results show that the adaptive threshold method proposed in this study is better than the fixed threshold for diagnosis, and the alarm times for the GENERATOR fault type, GENERATOR_BEARING fault type, and TRANSFORMER fault type are 1.5 h, 5.8 h, and 3 h earlier, respectively. [ABSTRACT FROM AUTHOR]

As the largest species of birds in Antarctica, penguins are called "biological indicators". Changes in the environment will cause population fluctuations. Therefore, developing a penguin census regularly will not only help carry out conservation activities but also provides a basis for studying climate change. Traditionally, scholars often use indirect methods, e.g., identifying penguin guano and establishing regression relationships to estimate the size of penguin colonies. In this paper, we explore the feasibility of automatic object detection algorithms based on aerial images, which locate each penguin directly. We build a dataset consisting of images taken at 400 m altitude over the island populated by Adelie penguins, which are cropped with a resolution of 640 × 640. To address the challenges of detecting minuscule penguins (often 10 pixels extent) amidst complex backgrounds in our dataset, we propose a new object detection network, named YoloPd (Yolo for penguin detection). Specifically, a multiple frequency features fusion module and a Bottleneck aggregation layer are proposed to strengthen feature representations for smaller penguins. Furthermore, the Transformer aggregation layer and efficient attention module are designed to capture global features with the aim of filtering out background interference. With respect to the latency/accuracy trade-off, YoloPd surpasses the classical detector Faster R-CNN by 8.5% in mean precision (mAP). It also beats the latest detector Yolov7 by 2.3% in F1 score with fewer parameters. Under YoloPd, the average counting accuracy reaches 94.6%, which is quite promising. The results demonstrate the potential of automatic detectors and provide a new direction for penguin counting. [ABSTRACT FROM AUTHOR]

The Three-River Headwaters Region (TRHR) is an important part of the ecological security barrier of the Qinghai–Tibet Plateau in China. Twenty years after the implementation of the TRHR ecological protection and construction project, the restoration degree and restoration potential of its major ecosystem services still lack clear quantification. In this paper, taking the core area of the nature reserve as the climax background of the TRHR zonal ecosystem, based on the multiple regression analysis (MLR) and model parameter control method based on the eco-geographical area, ecosystem types, and climate factors; the climax background, restoration degree, and restoration potential of TRHR's water retention (WR), soil retention (SR), and windbreak and sand fixation (WD) services were quantitatively researched. The main conclusions were as follows: (1) The evaluation method of climax background, restoration degree, and restoration potential based on fractional vegetation cover (FVC) can accurately quantify the regional differences of the restoration degree and restoration potential of TRHR's key ecosystem-regulating services. The restoration degree and restoration potential of WR and SR services showed a spatial pattern of high in the southeast and low in the northwest, and the restoration degree and restoration potential of WD services showed a spatial pattern of high in the west and low in the east, which was closely related to natural conditions such as precipitation and wind speed. (2) The proportion of restoration potential to climax background for WR, SR, and WD services were 48.38%, 62.15%, and 56.37%, respectively. (3) The implementation of the TRHR ecological project in the future should focus on the vicinity of the 400 mm dry and wet zone dividing line, as well as in the southeastern mountains, hills, and river valleys, to carry out degraded vegetation restoration and soil and water conservation measures to improve ecosystem services. Near-natural restoration measures should be considered in Zhiduo and Geermu in the western part of the TRHR, where wind erosion is high, and the restoration goals of ecological projects should be formulated in combination with local climatic conditions and restoration potential. [ABSTRACT FROM AUTHOR]

Objective: Precise classification of mass-forming intrahepatic cholangiocarcinoma (MF-ICC) and hepatocellular carcinoma (HCC) based on magnetic resonance imaging (MRI) is crucial for personalized treatment strategy. The purpose of the present study was to differentiate MF-ICC from HCC applying a novel deep-learning-based workflow with stronger feature extraction ability and fusion capability to improve the classification performance of deep learning on small datasets. Methods: To retain more effective lesion features, we propose a preprocessing method called semi-segmented preprocessing (Semi-SP) to select the region of interest (ROI). Then, the ROIs were sent to the strided feature fusion residual network (SFFNet) for training and classification. The SFFNet model is composed of three parts: the multilayer feature fusion module (MFF) was proposed to extract discriminative features of MF-ICC/HCC and integrate features of different levels; a new stationary residual block (SRB) was proposed to solve the problem of information loss and network instability during training; the attention mechanism convolutional block attention module (CBAM) was adopted in the middle layer of the network to extract the correlation of multi-spatial feature information, so as to filter the irrelevant feature information in pixels. Results: The SFFNet model achieved an overall accuracy of 92.26% and an AUC of 0.9680, with high sensitivity (86.21%) and specificity (94.70%) for MF-ICC. Conclusion: In this paper, we proposed a specifically designed Semi-SP method and SFFNet model to differentiate MF-ICC from HCC. This workflow achieves good MF-ICC/HCC classification performance due to stronger feature extraction and fusion capabilities, which provide complementary information for personalized treatment strategy. [ABSTRACT FROM AUTHOR]

Aiming at the problems of heavy workload of medical staff in the process of venous thrombosis prevention and treatment, error evaluation, missed evaluation, and inconsistent evaluation, we propose a joint extraction model of Chinese electronic medical records based on deep learning. The approach was to first construct the handshake annotation, then use bidirectional encoder representations from transformers (BERT) as the word vector embedding, then use the bidirectional long short-term memory network (BiLSTM) to extract the contextual features, and then integrate the contextual information into the process of normalizing the word vector. Experiments show that our proposed method achieves 93.3% and 94.3% of entity and relation F1 on the constructed electronic medical record dataset, which effectively improves the effect of medical information extraction. At the same time, the venous thromboembolism (VTE) risk factors extracted from the electronic medical record were used to judge the risk factors of venous thrombosis by means of rule reasoning. Compared with the assessment of clinicians on the Wells and Geneva scales, the accuracy rates of 84.7% and 86.1% were obtained. [ABSTRACT FROM AUTHOR]

The issue of identity authentication for online medical services has been one of the key focuses of the healthcare industry in recent years. Most healthcare organizations use centralized identity management systems (IDMs), which not only limit the interoperability of patient identities between institutions of healthcare, but also create isolation between data islands. The more important matter is that centralized IDMs may lead to privacy disclosure. Therefore, we propose Health-zkIDM, a decentralized identity authentication system based on zero-knowledge proof and blockchain technology, which allows patients to identify and verify their identities transparently and safely in different health fields and promotes the interaction between IDM providers and patients. The users in Health-zkIDM are uniquely identified by one ID registered. The zero-knowledge proof technology is deployed on the client, which provides the user with a proof of identity information and automatically verifies the user's identity after registration. We implemented chaincodes on the Fabric, including the upload of proof of identity information, identification, and verification functions. The experiences show that the performance of the Health-zkIDM system can achieve throughputs higher than 400 TPS in Caliper. [ABSTRACT FROM AUTHOR]

The gut microbiome is related to many major human diseases, and it is of great significance to study the structure of the gut microbiome under different conditions. Multivariate statistics or pattern recognition methods were often used to identify different structural patterns in gut microbiome data. However, these methods have some limitations. Minimal hepatic encephalopathy (MHE) datasets were taken as an example. Due to the physical lack or insufficient sampling of the gut microbiome in the sequencing process, the microbiome data contains many zeros. Therefore, the geometric mean of pairwise ratios (GMPR) was used to normalize gut microbiome data, then Spectrum was used to analyze the structure of the gut microbiome, and lastly, the structure of core microflora was compared with Network analysis. GMPR calculates the Intraclass correlation coefficient (ICC), whose reproducibility was significantly better than other normalization methods. In addition, running-time, Normalized Mutual Information (NMI), Davies-Boulding Index (DBI), and Calinski-Harabasz index (CH) of GMPR+Spectrum were far superior to other clustering algorithms such as M3C, iClusterPlus. GMPR+Spectrum can not only perform better but also effectively identify the structural differences of intestinal microbiota in different patients and excavate the unique critical bacteria such as Akkermansia, and Lactobacillus in MHE patients, which may provide a new reference for the study of the gut microbiome in disease. [ABSTRACT FROM AUTHOR]

Adversarial attacks deceive deep neural network models by adding imperceptibly small but well-designed attack data to the model input. Those attacks cause serious problems. Various defense methods have been provided to defend against those attacks by: (1) providing adversarial training according to specific attacks; (2) denoising the input data; (3) preprocessing the input data; and (4) adding noise to various layers of models. Here we provide a simple but effective Noise-Fusion Method (NFM) to defend adversarial attacks against DNN image classification models. Without knowing any details about attacks or models, NFM not only adds noise to the model input at run time, but also to the training data at training time. Two l∞-attacks, the Fast Gradient Signed Method (FGSM) and the Projected Gradient Descent (PGD), and one l1-attack, the Sparse L1 Descent (SLD), are applied to evaluate defense effects of the NFM on various deep neural network models which used MNIST and CIFAR-10 datasets. Various amplitude noises with different statistical distribution are applied to show the defense effects of the NFM in different noise. The NFM also compares with an adversarial training method on MNIST and CIFAR-10 datasets. Results show that adding noise to the input images and the training images not only defends against all three adversarial attacks but also improves robustness of corresponding models. The results indicate possibly generalized defense effects of the NFM which can extend to other adversarial attacks. It also shows potential application of the NFM to models not only with image input but also with voice or audio input. [ABSTRACT FROM AUTHOR]

Acute urethral injuries caused by urethral endoscopy and other mechanical injuries are the main reasons for secondary infection and late urethral stricture. However, there are no studies to explore the transcriptomic changes in urethral injury and the molecular mechanism of urethral injury, which is important for the treatment and cure of urethral injury. Therefore, we used RNA-seq and sRNA-seq profiles from normal and injured urethral tissues to identify and characterize differentially expressed mRNAs and miRNAs. In total, we found 166 differentially expressed mRNAs, of which 69 were upregulated, and 97 were downregulated in injured urethral tissues. The differentially expressed mRNAs were mainly involved in the positive regulation of epithelial cell differentiation, focal adhesion, cell adhesion molecules, protein activation cascade, complement activation, complement and coagulation cascades, and chemokine-mediated signaling pathway. Additionally, we found six upregulated and four downregulated miRNAs, respectively, in the injured urethral tissues. Notably, their target genes were involved in the vascular endothelial growth factor receptor 2 binding, PI3k-Akt signaling pathway, and Notch signaling pathway. In summary, our results suggest that the cell damage response induced by mechanical injury activates the pathological immune response in a variety of ways in injured urethral tissues. [ABSTRACT FROM AUTHOR]

Located at the junction between the continental climate region and marine climate region, the Qilian Mountains have experienced significant climate change. Vegetation phenology in the Qilian Mountains is sensitive to climate change. However, the response of vegetation phenology to temperature and precipitation change is still unclear, and the same is true for their interactions. First, we extracted grassland phenological parameters such as SOS (the start of the growing season), EOS (the end of the growing season), and LOS (the length of the growing season) from revised MODIS-NDVI data in the Qilian Mountains during the period from 2000 to 2019. Second, we analyzed change trends of the phenological parameters, temperature, and precipitation. Furthermore, the effects of each meteorological element changes and their interaction on multiple phenological parameters were detected using the GeoDetector method. The result implied that (1) the SOS in most areas except the northwestern mountain region showed an advanced trend (10 d/10a); the EOS showed a delayed trend in the southeast (5 d/10a), and an advanced trend (5 d/10a) in the northwest; the LOS showed an extended trend (10 d/10a) in the southeast, and a shortened trend (5 d/10a) in the northwest. (2) Compared with a single meteorological element in a single period, the interaction of temperature and precipitation in different periods had a higher impact on grassland phenology, with the maximum q-value increasing by about 0.4 for each phenological parameter. (3) The change in the grassland phenology in the Qilian Mountains was inconsistently complete with climate change in the spatial distribution. Our research reveals the response of grassland phenology to the interaction of different meteorological elements in different periods. Compared with a single element, this can reflect the response of vegetation phenology to climate change more comprehensively. [ABSTRACT FROM AUTHOR]

Kongsfjorden is a small Arctic fjord but with great hydrographic complexity and has changed greatly due to the climate change. Arctic warming has increased melts of sea ice and glaciers that results in higher freshwater content. Microbial community variability and increasing terrestrial input were detected continuously in recent years ITag eukaryotic 18S rRNA V4 metabarcoding, photosynthetic pigments analysis and epifluorescence microscopy were used to reveal the dominant species of small eukaryotic community (<20 μm). Both Spearman correlation and redundant analysis were used to study the correlation between the small eukaryotes and the environmental conditions. In the present study, the surface water with salinity lower than 34 was thicker than in summers of previous years. The freshwater mixotrophic chrysophyte Poterioochromonas malhamensis was found for the first time as the dominant species. No general trends were found for the contributions of P. malhamensis to the total reads of small eukaryotes in water depths; and no obvious differences were found at different stations and water masses. Phagotrophy, which is more common than phototrophy at all times in P. malhamensis, is thought to be the main reason for the prevalence of P. malhamensis in Kongsfjorden. The occurrence of P. malhamensis induced a disorder in the small eukaryotic community, which biodiversity and composition showed weak correlation with the water masses. The dominance of the freshwater-originating phytoplankton may indicate an ecosystem change in the Kongsjforden, which probably might become more remarkable in the future as the climate continues to change. [ABSTRACT FROM AUTHOR]

Objective: Precise classification of mass-forming intrahepatic cholangiocarcinoma (MF-ICC) and hepatocellular carcinoma (HCC) based on magnetic resonance imaging (MRI) is crucial for personalized treatment strategy. The purpose of the present study was to differentiate MF-ICC from HCC applying a novel deep-learning-based workflow with stronger feature extraction ability and fusion capability to improve the classification performance of deep learning on small datasets., Methods: To retain more effective lesion features, we propose a preprocessing method called semi-segmented preprocessing (Semi-SP) to select the region of interest (ROI). Then, the ROIs were sent to the strided feature fusion residual network (SFFNet) for training and classification. The SFFNet model is composed of three parts: the multilayer feature fusion module (MFF) was proposed to extract discriminative features of MF-ICC/HCC and integrate features of different levels; a new stationary residual block (SRB) was proposed to solve the problem of information loss and network instability during training; the attention mechanism convolutional block attention module (CBAM) was adopted in the middle layer of the network to extract the correlation of multi-spatial feature information, so as to filter the irrelevant feature information in pixels., Results: The SFFNet model achieved an overall accuracy of 92.26% and an AUC of 0.9680, with high sensitivity (86.21%) and specificity (94.70%) for MF-ICC., Conclusion: In this paper, we proposed a specifically designed Semi-SP method and SFFNet model to differentiate MF-ICC from HCC. This workflow achieves good MF-ICC/HCC classification performance due to stronger feature extraction and fusion capabilities, which provide complementary information for personalized treatment strategy.