Biomimetic sensors offer a rapid route for estimating active constituent contents in Chinese medicinal materials, but their practical use at procurement sites and production workshops is constrained by small and skewed datasets, repeated task-specific model configuration, and the limited operational traceability of many nonlinear models. This study proposes the Traditional Chinese Medicine Unified Sensor-based Prediction (TCM-USP) framework for traceable preliminary screening rather than confirmatory laboratory quantification. The same modeling workflow was applied across different medicinal material–sensor combinations and integrates controlled symbolic feature construction with density-aware robust partial least squares modeling. The framework was evaluated on six prediction tasks involving five medicinal materials and three sensor types. TCM-USP achieved higher R2 values than raw-feature PLS in all six tasks and achieved the highest R2 among the evaluated models in four tasks. Although SVR or GPR achieved higher R2 values in the remaining two tasks, TCM-USP generated compact prediction formulas directly expressed in terms of the original sensor readings, enabling independent calculation, audit, and rapid batch-level decision support. These results support the feasibility of TCM-USP for traceable preliminary screening across the investigated small-sample and low-dimensional sensor tasks, while pharmacopoeial methods remain necessary for confirmatory quantification.

Missing data in morphological trait datasets pose a persistent challenge to ecological and evolutionary research, frequently compromising model inference and predictive accuracy. We propose THORBFNN, a three-stage hybrid imputation framework that integrates regularized K-means clustering, Radial Basis Function Neural Networks (RBFNNs), and hierarchical Bayesian optimization to accurately recover missing avian morphological traits. The framework partitions species into clusters using regularized K-means, enhancing the preservation of local morphological structure through inter-cluster separation. Within each cluster, RBFNNs model nonlinear dependencies among traits using input features selected by Pearson correlation with the target trait. Key hyperparameters such as the number of clusters and RBF width are optimized via hierarchical Bayesian optimization to balance generalization and model complexity. When applied to a global avian trait dataset comprising over 10,000 individuals and 11 morphological traits, THORBFNN outperforms K-nearest neighbors and Random Forest imputation across four focal traits, achieving higher R 2 and lower errors (THORBFNN: R 2 = 0.9003, RMSE = 0.1652, MAE = 0.1096; KNN: R 2 = 0.8864, RMSE = 0.1668, MAE = 0.1248; Random Forest: R 2 = 0.8573, RMSE = 0.2134, MAE = 0.1584). Ablation experiments comparing models trained on complete cases versus mean-imputed data confirm that THORBFNN captures genuine trait covariation rather than statistical artifacts. THORBFNN requires no phylogenetic information and scales efficiently to datasets with thousands of individuals, offering a practical pathway for integrating machine learning into biodiversity trait analysis.

In this study, we introduce Dual-Branch BioTraitNet, a deep-learning model tailored for trait imputation in small-sample ecological and biological datasets. By combining unsupervised and supervised learning strategies, the model jointly leverages quantitative and qualitative trait information. Its dual-branch architecture enables efficient learning under data-sparse conditions and generalizes well across diverse taxa. On the lizard dataset, the model achieved R2 values of 0.862 for mean body length and 0.67 for average body weight; on the fish dataset, R2 values for maximum body length, minimum spawning temperature, and egg diameter were 0.876, 0.402, and 0.496, respectively. Unlike conventional approaches such as K-nearest neighbors (KNN) and genetic algorithms (and their variants), which are often prone to overfitting or underfitting, BioTraitNet demonstrates strong predictive stability and robustness. This is evident in its consistent avoidance of negative R2 values. Notably, it maintains high accuracy even without incorporating phylogenetic information, making it particularly suitable for scenarios where evolutionary data are missing or uncertain. The proposed framework offers a flexible and reliable solution for addressing missing trait data in ecological and evolutionary research. The computational Python code was available from https://github.com/BB-yu/Dual-Branch-BioTraitNet.

Accurate segmentation of skin lesions is crucial for reliable clinical diagnosis and effective treatment planning. Automated techniques for skin lesion segmentation assist dermatologists in early detection and ongoing monitoring of various skin diseases, ultimately improving patient outcomes and reducing healthcare costs. To address limitations in existing approaches, we introduce a novel U-shaped segmentation architecture based on our Residual Space State Block. This efficient model, termed ‘SSR-UNet,’ leverages bidirectional scanning to capture both global and local features in image data, achieving strong performance with low computational complexity. Traditional CNNs struggle with long-range dependencies, while Transformers, though excellent at global feature extraction, are computationally intensive and require large amounts of data. Our SSR-UNet model overcomes these challenges by efficiently balancing computational load and feature extraction capabilities. Additionally, we introduce a spatially-constrained loss function that mitigates gradient stability issues by considering the distance between label and prediction boundaries. We rigorously evaluated SSR-UNet on the ISIC2017 and ISIC2018 skin lesion segmentation benchmarks. The results showed that the accuracy of Mean Intersection Over Union, Classification Accuracy and Specificity indexes in ISIC2017 datasets reached 80.98, 96.50 and 98.04, respectively, exceeding the best indexes of other models by 0.83, 0.99 and 0.38, respectively. The accuracy of Mean Intersection Over Union, Dice Coefficient, Classification Accuracy and Sensitivity on ISIC2018 datasets reached 82.17, 90.21, 95.34 and 88.49, respectively, exceeding the best indicators of other models by 1.71, 0.27, 0.65 and 0.04, respectively. It can be seen that SSR-UNet model has excellent performance in most aspects.

QuestionsAccurately measuring species distribution patterns has long been a fundamental task in community and spatial ecology. This study tries to address the following ecological problems: (1) Are distance-based methods powerful and sensitive enough to detect subtle changes of spatial distributional patterns of species in a distributional map, particularly when it comes to complete random distribution? (2) How can spatial distributional aggregation patterns be effectively and rapidly compared when the size of spatial distributional data is large?LocationGlobal.MethodsWe propose a one-dimensional aggregation index, quadrat nearest neighboring distance (QNN), integrated with an optimal spatial subdivision protocol. This method was evaluated via numerical simulations and applied to empirical species distribution data and compared against Clark and Evans' R index, which is built upon two-dimensional nearest neighboring distance (NND).ResultsQNN is accurate in detecting complete random distribution over different population sizes of species. The accuracy of QNN depended on using an optimal gridding size. The accuracy of detecting the complete random distribution of species is guaranteed when the sampling grain size in QNN is around the optimal gridding size. Furthermore, QNN showed high computational efficiency in extensive numerical simulations.ConclusionsQNN offers a powerful, efficient, and scalable tool for detecting and comparing species spatial distribution patterns. Its enhanced sensitivity to random distributions and dependence on optimal gridding make it especially suitable for analyzing large-scale or high-resolution spatial data.

High-quality images can provide consumers with more product information, enhance purchase intention, and further affect the brand's economic and financial benefits. How to improve image quality has become a key challenge for online product presentations. As one of the most popular image processing tools, image super-resolution not only improves image quality with more details but also saves image storage space and transmission bandwidth. Therefore, in this study, we first review the relationship between image quality and consumer purchase intention. Based on the characteristics of product images containing multiple images with similar content, we then propose a novel reference-based super resolution method based on Fourier transform. The proposed Frequency-Spatial Aggregation block extracts complementary spatial and frequency features to activate more useful pixels. Experiments show that our method achieves state-of-the-art results with fewer parameters and faster running time. In addition, in order to provide a comprehensive evaluation of our method in the context of product images, we collect image pairs from e-commerce platforms and contribute a new dataset named PRSR dataset.

RATIONALE AND OBJECTIVES:Cardiac magnetic resonance imaging is a crucial tool for analyzing, diagnosing, and formulating treatment plans for cardiovascular diseases. Currently, there is very little research focused on balancing cardiac segmentation performance with lightweight methods. Despite the existence of numerous efficient image segmentation algorithms, they primarily rely on complex and computationally intensive network models, making it challenging to implement them on resource-constrained medical devices. Furthermore, simplified models designed to meet the requirements of device lightweighting may have limitations in comprehending and utilizing both global and local information for cardiac segmentation. MATERIALS AND METHODS:We propose a novel 3D high-performance lightweight medical image segmentation network, HL-UNet, for application in cardiac image segmentation. Specifically, in HL-UNet, we propose a novel residual-enhanced Adaptive attention (REAA) module that combines residual-enhanced connectivity with an adaptive attention mechanism to efficiently capture key features of input images and optimize their representation capabilities, and integrates the Visual Mamba (VSS) module to enhance the performance of HL-UNet. RESULTS:Compared to large-scale models such as TransUNet, HL-UNet increased the Dice of the right ventricular cavity (RV), left ventricular myocardia (MYO), and left ventricular cavity (LV), the key indicators of cardiac image segmentation, by 1.61%, 5.03% and 0.19%, respectively. At the same time, the Params and FLOPs of the model decreased by 41.3 M and 31.05 G, respectively. Furthermore, compared to lightweight models such as the MISSFormer, the HL-UNet improves the Dice of RV, MYO, and LV by 4.11%, 3.82%, and 4.33%, respectively, when the number of parameters and computational complexity are close to or even lower. CONCLUSION:The proposed HL-UNet model captures local details and edge information in images while being lightweight. Experimental results show that compared with large-scale models, HL-UNet significantly reduces the number of parameters and computational complexity while maintaining performance, thereby increasing frames per second (FPS). Compared to lightweight models, HL-UNet shows substantial improvements across various key metrics, with parameter count and computational complexity approaching or even lower.

The novel coronavirus disease — COVID-19 is a historic catastrophe that has caused many devastating impacts on human life and wellness. Researchers in academia and industry strive to understand the causes of this pandemic disease and find new therapeutics combating it. Consequently, the number of COVID-19 related publications increases rapidly, and it is too difficult for medical researchers and practitioners to keep up with the latest research and development. Literature filtering and categorization, and knowledge discovery can use text mining as a powerful tool. In this paper, we propose a text mining method to explore the categories of COVID-19 related themes and identify the standard methodologies that have been used. We discuss the potential limitations of this preliminary study and present future perspectives related to COVID-19 research. This paper provides an quantitative and qualitative mixed analysis example of using some research papers by data mining method to dig out several hidden information and set up a foundation for data scientists to develop more effective algorithms to deal with COVID-19 related problems.

The distribution of species is not random in space. At the finest-resolution spatial scale, that is, field sampling locations, distributional aggregation level of different species would be determined by various factors, for example spatial autocorrelation or environmental filtering. However, few studies have quantitatively measured the importance of these factors. In this study, inspired by the statistical properties of a Markov transition model, we propose a novel additive framework to partition local multispecies distributional aggregation levels for sequential sampling-derived field biodiversity data. The framework partitions the spatial distributional aggregation of different species into two independent components: regional abundance variability and the local spatial inertia effect. Empirical studies from field amphibian surveys through line-transect sampling in southwestern China (Minya Konka) and central-southern Vietnam showed that local spatial inertia was always the dominant mechanism structuring the local occurrence and distributional aggregation of amphibians in the two regions with a latitudinal gradient from 1200 to nearly 4000 m. However, regional abundance variability is still nonnegligible in highly diverse tropical regions (i.e. Vietnam) where the altitude is not higher than 2000 m. In summary, we propose a novel framework that shows that the multispecies distributional aggregation level can be structured by two additive components. The two partitioned components could be theoretically independent. These findings are expected to deepen our understanding of the local community structure from the perspective of both spatial distribution and regional diversity patterns. The partitioning framework might have potential applications in field ecology and macroecology research. Innovative study proposes an additive partitioning framework for multispecies distributional aggregation, integrating regional abundance variability and local spatial inertia. Empirical validation conducted in southwestern China and central-southern Vietnam. Offers valuable insights for ecologists.image

Malicious websites present a substantial threat to the security and privacy of individuals using the internet. Traditional approaches for identifying these malicious sites have struggled to keep pace with evolving attack strategies. In recent years, language models have emerged as a potential solution for effectively detecting and categorizing malicious websites. This study introduces a novel Bidirectional Encoder Representations from Transformers (BERT) model, based on the Transformer encoder architecture, designed to capture pertinent characteristics of malicious web addresses. Additionally, large-scale language models are employed for training, dataset assessment, and interpretability analysis. The evaluation results demonstrate the effectiveness of the large language model in accurately classifying malicious websites, achieving an impressive precision rate of 94.42%. This performance surpasses that of existing language models. Furthermore, the interpretability analysis sheds light on the decision-making process of the model, enhancing our understanding of its classification outcomes. In conclusion, the proposed BERT model, built on the Transformer encoder architecture, exhibits robust performance and interpretability in the identification of malicious websites. It holds promise as a solution to bolster the security of network users and mitigate the risks associated with malicious online activities.

Electric power operation, as one of the key fields in the world, faces particularly prominent safety issues. Ensuring the safety of operators has become the most fundamental requirement in power operation. However, there are some safety hazards in power construction. These hazards are mainly due to weak safety awareness among staff and the failure to standardize the wearing of safety helmets. In order to effectively address this situation, technical means such as video surveillance technology and computer vision technology can be utilized to monitor whether staff are wearing helmets and provide timely feedback. Such measures will greatly enhance the safety level of power operation. This paper proposes an improved lightweight helmet detection algorithm named YOLO-M3C. The algorithm first replaces the YOLOv5s backbone network with MobileNetV3, successfully reducing the model size from 13.7 MB to 10.2 MB, thereby increasing the model’s detection speed from 42.0 frames per second to 55.6 frames per second. Then, the CA attention mechanism is introduced into the backbone network to enhance the feature extraction capability of the model. Finally, in order to further improve the detection recall rate and accuracy of the model, a knowledge distillation of the model was carried out. The experimental results show that, compared with the original YOLOv5s algorithm, the average accuracy of the improved YOLO-M3C algorithm is improved by 0.123, and the recall rate is the same. These results verify that the algorithm YOLO-M3C has excellent performance in target detection and recognition, which can improve accuracy and confidence, while reducing false detection and missing detection, and effectively meet the needs of helmet-wearing detection.

Conventional wound dressings fail to satisfy the requirements and needs of wounds in various stages. It is challenging to develop a multifunctional dressing that is hemostatic, antibacterial, anti-inflammatory, and promotes wound healing. Therefore, this study aimed to develop a multifunctional sponge dressing for the full-stage wound healing based on copper and two natural products, Bletilla striata polysaccharide (BSP) and peony leaf extract (PLE). The developed BSP-Cu-PLE sponges were characterized by SEM, XRD, FTIR, and XPS to assess micromorphology and elemental composition. Their properties and bioactivities were also verified by the further experiments, whereby the findings revealed that the BSP-Cu-PLE sponges had improved water absorption and porosity while exhibiting excellent antioxidative, biocompatible, and biodegradable properties. Moreover, the antibacterial test revealed that BSP-Cu-PLE sponges had superior antibacterial activity against S. aureus and E. coli. Furthermore, the hemostatic activity of BSP-Cu-PLE sponges was significantly enhanced in a rat liver trauma model. Most notably, further studies have demonstrated that the BSP-Cu-PLE sponges could significantly (p < 0.05) accelerate the healing process of skin wounds by stimulating collagen deposition, promoting angiogenesis, and decreasing inflammatory cells. In summary, the BSP-Cu-PLE sponges could provide a new strategy for application in clinical setting for full-stage wound healing.

This paper proposes a robust adaptive filter based on the exponent sin cost to improve the capability against Gaussian or multiple types of non-Gaussian noises of the adaptive filtering algorithm when dealing with time-varying/time-invariant linear systems function exponent sin(ExpSin).Then a variable step-size(VSS)-ExpSin algorithm is extended further.Besides,the stepsize,the convergence,and the steady-state performance of the proposed algorithm are validated experimentally.The Monte Carlo simulation results of linear system identification illustrate the principle and efficiency of this proposed adaptive filtering algorithm.Results suggest that the proposed adaptive filtering algorithm has superior performance when estimating the unknown linear systems under multiple-types measurement noises.

Smart cities have been a popular topic for the city stakeholders. A smart city is the next urban lifestyle that citizens expect. Due to the hypercompetitive and globalized economy, many cities have already started or are about to start their smart city projects. There is no uniform benchmark to evaluate the smart cities’ performance. Several organizations use their own indicators to evaluate smart cities worldwide or nationwide. This research paper leverages fuzzy logic to label smart city leaders and followers based on various organization’s evaluation meta results and then uses machine learning techniques to identify the key characteristics of leaders and followers. Based on the training data performance, the Support Vector Machine (SVM) is used to predict who will be the next smart city leader or follower. According to the proposed prediction framework, we have successfully predicted 30 smart city leaders and 20 followers.

Objective:To analyze the epidemiological characteristics of clustered outbreaks in the early stage of the coronavirus disease 2019 (COVID-19) epidemic in Haidian district, Beijing, so as to provide reference for future epidemic prevention and control.Methods:The data of patients in the COVID-19 clustered outbreaks in Haidian district from January 20 to February 28, 2020 were collected, and the characteristics of the epidemic were analyzed by descriptive epidemiological methods.Results:In the early stage of the epidemic, a total of 13 clustered outbreaks occurred in Haidian District, involving 45 infected patients, including 5 deaths. The outbreaks were mainly family clusters (11/13) and imported cases from other provinces (9/13). Four outbreaks caused by transmission during the incubation period were observed. The incubation period, M ( P25, P75) was 5 (3.5, 7.5) d and the serial intervals, M ( P25, P75) was 5 (3.5, 11.5) d. For patients proactively seeking medical attention, 50% (13/26) had multiple visits at hospitals. The nucleic acid of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was detected in 84.1% (37/44) of the cases between 2 days before and 7 days after disease onset. The overall secondary attack rate was 11.6% (29/249) while the rate in family was 32.1% (27/84). Conclusions:In the early stage of the COVID-19 epidemic in Haidian district, the outbreaks were mainly family clusters. Improving the ability in early identification of the cases, quarantine of close contacts and timely sample collection will assist controlling the epidemic from expanding.

Natural scene text detection has an important role to play in getting textual information from natural scenes. With the continuous development of deep learning, natural scene text detection methods are emerging and achieving better results on detection tasks. In this paper, analysis, and summary of the current stage of deep learning-based text algorithms for natural scenes, can be divided into two types: region of the proposal and semantic segmentation, and the content of these two series of related algorithms is described. Secondly, a publicly available dataset and detection performance metrics for scene text detection are presented. Ultimately, the research in scene text detection is summarized and looked forward to in the hope of providing new research directions for subsequent algorithms.

Environmental protection is still a key issue that cannot be ignored at this stage of social development. With the development of artificial intelligence, various technologies increasingly tend to be widely used in the field of environmental protection, such as searching the wilderness through an unmanned aerial vehicle (UAV) and cleaning garbage by robots. Traditional object detection algorithms for this scenario suffer from low accuracy and high computational cost. Therefore, this paper proposes an algorithm applied to automatic garbage detection and instance segmentation in complex scenes. First, we construct sample-fused feature pyramid networks (SF-FPN) to achieve multi-scale feature sampling on multiple levels, to enhance the semantic representation of features. Second, adding the mask branch based on conditional convolution, introducing the idea of instance-filters to automatically generate the filter parameters of the Fully Convolutional Networks (FCN), to realize the instance-level pixel classification. Moreover, the Atrous Spatial Pyramid Pooling (ASPP) module is introduced to encode the feature information in a dense way to assist the generation of MASK. Finally, the object is detected and the instance is segmented by a two-branch structure. In addition, we also perform data augmentation on the original dataset to prevent model overfitting. The proposed algorithm reaches 82.7 and 72.4 according to the mAP index of detection and instance segmentation while using the public TACO dataset.

目的 分析2020年至2021年北京市海淀区HIV抗体筛查试验结果与确证试验结果的相关性,为提高实验室的检测能力提供技术支持.方法 对筛查实验室送检的1420例HIV抗体筛查有反应样本采用免疫印迹法(WB)进行确证试验,将筛查试验结果按不同的人群、检测方法、S/CO值与确证结果进行对比分析.结果 1420例HIV抗体筛查有反应样本的确证阳性率为34.72％,其中男性(49.58％)确证阳性率明显高于女性(3.90％),21～40岁年龄组确证阳性率最高(39.95％),来自海淀区疾病预防控制中心仁爱社区送检样本的确证阳性率最高(97.58％),最低为血液中心(23.09％).不同筛查方法之间的确证阳性率差异有统计学意义(P＜0.05),快速检测法最高(94.81％),ELISA法最低(32.17％).当ELISA法S/CO＞10或CLIA法S/CO＞50或ECLIA法S/CO＞50时,确证阳性率能达到较高的水平.结论 应加强对筛查实验室的质量控制,提高HIV抗体检测水平.筛查试验的S/CO值与确证试验结果具有一定的相关性,可在一定程度上通过S/CO比值预测确证试验结果.

With the development of deep learning technology, handwritten character recognition has become the basic hotspot of research. At the same time, the recognition of handwriting in minority languages is gradually worthy of attention. In this paper, a robust and novel improved deep convolutional neural network method is proposed to recognize normal digital handwriting images and minority digital handwritten images (Yi language). To improve the recognition accuracy of handwritten fonts, there are four improvements in our work. Firstly, separable convolution is adopted in the network to cope with the problem of large parameters. Secondly, data augmentation techniques (Gaussian noise, color perturbation, and random rotation) are utilized to improve the expression ability of the dataset and the generalization capabilities of the model. Finally, the shortcut block and the weight initialization are introduced in our network for the problems of network degradation and gradient saturation. Based on deep learning technology, this paper studies the classification and recognition of handwritten digits, and improves the accuracy of the recognition in numbers such as invoices in financial bills. The recognition accuracy of 99.82% has been achieved on MNIST dataset, and 98.79% on the self-building dataset of Yi handwritten digits (YHDD).

Text sentiment tendency analysis is a hot task in natural language processing. And text as the essential expression form of language, both individual word information, and overall utterance, deserves to be focused on. This paper proposes a fusion model to achieve high precision text sentiment analysis. This model combines the advantages of CNN to extract local information of text and BiLSTM to extract contextual association of text and introduces the attention mechanism to increase the focus on words with a solid emotional tendency in the text. The training datasets are comments that crawled from several social media sites such as Facebook, Twitter, Instagram, WhatsApp, etc. Based on the attention mechanism, this paper investigates the semantic sentiment analysis to reach the study of classification prediction for analyzing the positive and negative sentiment of financial news, social media, etc. The experimental results show that the proposed method can better extract features from the text and classify them than other baseline models.