Explore the vast range of titles available.

In the context of rapid urbanization, PM2.5 pollution poses a significant threat to public health and environmental quality. Current spatiotemporal analysis methods often lack sufficient accuracy. To address this, this study uses spatiotemporal analysis and Spatial Temporal Graph Convolutional Networks (ST-GCN) to evaluate the variation and driving factors of PM2.5 concentrations in the Beijing-Tianjin-Hebei (BTH) urban agglomeration from 2014 to 2024, and to make predictions. ST-GCN effectively captures spatial features and temporal dependencies of PM2.5, revealing intercity relationships more accurately. SHapley Additive exPlanations (SHAP) provide model interpretability, identifying key factors influencing PM2.5 concentrations. The results from experiments on the BTH dataset are as follows: (1) The proposed algorithm accurately captures the spatiotemporal characteristics of pollution in the BTH urban agglomeration, with an R2 of 0.96. (2) PM2.5 concentrations in the BTH urban agglomeration decreased significantly from 2014 to 2024, with a cumulative reduction of 60.00%, especially in the southwestern areas, though annual averages still exceeded national standards. (3) SHAP analysis shows that PM10 and CO are the main positive contributors to PM2.5 concentrations. (4) From 2025 to 2030, PM2.5 concentrations are expected to decline, with a 2030 concentration of 34.47 μg m−3.

Green citrus detection in citrus orchards provides reliable support for production management chains, such as fruit thinning, sunburn prevention and yield estimation. In this paper, we proposed a lightweight object detection YOLOv5-CS (Citrus Sort) model to realize object detection and the accurate counting of green citrus in the natural environment. First, we employ image rotation codes to improve the generalization ability of the model. Second, in the backbone, a convolutional layer is replaced by a convolutional block attention module, and a detection layer is embedded to improve the detection accuracy of the little citrus. Third, both the loss function CIoU (Complete Intersection over Union) and cosine annealing algorithm are used to get the better training effect of the model. Finally, our model is migrated and deployed to the AI (Artificial Intelligence) edge system. Furthermore, we apply the scene segmentation method using the “virtual region” to achieve accurate counting of the green citrus, thereby forming an embedded system of green citrus counting by edge computing. The results show that the mAP@.5 of the YOLOv5-CS model for green citrus was 98.23%, and the recall is 97.66%. The inference speed of YOLOv5-CS detecting a picture on the server is 0.017 s, and the inference speed on Nvidia Jetson Xavier NX is 0.037 s. The detection and counting frame rate of the AI edge system-side counting system is 28 FPS, which meets the counting requirements of green citrus.

With the development of agricultural information technology, the Internet of Things and blockchain have become important in the traceability of agricultural products. Sensors collect real-time data in agricultural production and a blockchain provides a secure and transparent storage medium for these data, which improves the transparency and credibility of agricultural product traceability. However, existing agricultural product traceability solutions are limited by the immutability of the blockchain, making it difficult to delete erroneous data and modify the scope of data sharing. This damages the credibility of traceability data and is not conducive to the exchange and sharing of information among enterprises. In this article, we propose an agricultural product traceability data management scheme based on a redactable blockchain. This scheme allows agricultural enterprises to encrypt data to protect privacy. In order to facilitate the maintenance and sharing of data, we introduce a chameleon hash function to provide data modification capabilities. Enterprises can fix erroneous data and update the access permissions of the data. To improve the efficiency of block editing, our scheme adopts a distributed block editing method. This method supports threshold editing operations, avoiding single-point-of-failure issues. We save records of data modifications on the blockchain and establish accountability mechanisms to identify malicious entities. Finally, in this paper we provide a security analysis of our proposed solution and verify its effectiveness through experiments. Compared with the existing scheme, the block generating speed is improved by 42% and the block editing speed is improved by 29.3% at 125 nodes.

Maritime objects frequently exhibit low-quality and insufficient feature information, particularly in complex maritime environments characterized by challenges such as small objects, waves, and reflections. This situation poses significant challenges to the development of reliable object detection including the strategies of loss function and the feature understanding capabilities in common YOLOv8 (You Only Look Once) detectors. Furthermore, the widespread adoption and unmanned operation of intelligent ships have generated increasing demands on the computational efficiency and cost of object detection hardware, necessitating the development of more lightweight network architectures. This study proposes the EL-YOLO (Efficient Lightweight You Only Look Once) algorithm based on YOLOv8, designed specifically for intelligent ship object detection. EL-YOLO incorporates novel features, including adequate wise IoU (AWIoU) for improved bounding box regression, shortcut multi-fuse neck (SMFN) for a comprehensive analysis of features, and greedy-driven filter pruning (GDFP) to achieve a streamlined and lightweight network design. The findings of this study demonstrate notable advancements in both detection accuracy and lightweight characteristics across diverse maritime scenarios. EL-YOLO exhibits superior performance in intelligent ship object detection using RGB cameras, showcasing a significant improvement compared to standard YOLOv8 models.

Transradial access (TRA) in neurointervention has gained popularity due to fewer procedural complications and reduced recovery time compared to transfemoral access (TFA). This study aimed to assess the feasibility and safety of using 0.018-inch guidewire (018’’GW)-supported distal access catheters (DAC) in establishing transradial neurointerventional access, in comparison to the Ballast long sheath. We performed a retrospective review of a prospective database of transradial neurointervention with Ballast long sheath-supported DACs or 018’’GW-supported DACs. Patient demographics, baseline clinical characteristics, and detailed procedural information were recorded. The success rate of TRA neurointervention using 018’’GW-supported DACs was comparable to that of the Ballast long sheath protocol (100% vs. 96.3%, p  = 0.19). Both protocols achieved comparable placement heights of DACs and outcomes for aneurysm and symptomatic intracranial atherosclerotic stenosis (sICAS) treatment. Moreover, the 018’’GW-supported DACs significantly decreased 6-month radial artery occlusion (RAO) rates (5.88% vs. 18.18%, p  = 0.045) without any major vascular or neurological complications. This study highlights the feasibility and safety of 018’’GW-supported DACs in TRA neurointervention, offering a viable alternative with reduced complications and enhanced distal stability in comparison with the Ballast long sheath.

Advanced computer vision techniques, particularly deep learning–based object detection, are enhancing the accuracy and efficiency of product identification in retail settings, driving the integration of intelligent systems within urban environments and smart cities. To address the high computational cost and slow detection speed of existing methods, this study proposes LSR-YOLO, a lightweight object detection framework based on YOLOv8n, designed for deployment in robots and intelligent devices. The model introduces architectural optimizations, including the CSPHet-CBAM attention module, to strengthen feature representation, followed by a channel pruning algorithm tailored to the new architecture to reduce redundancy while maintaining accuracy. Experiments on the Locount dataset demonstrate that LSR-YOLO achieves an inference speed of 357.1 FPS with mAP50 of 72.2% and mAP50-95 of 47.8%. Compared with the baseline YOLOv8n, LSR-YOLO increases inference speed by 246.7 FPS, making it substantially faster and more suitable for real-time retail applications. With only 2,114,768 parameters and 6.6 GFLOPs, it is also significantly lighter than advanced models such as YOLOv11. Furthermore, validation on the COCO dataset confirms the model’s superior generalization ability, underscoring its advantages in both accuracy and computational efficiency.

As maximum power point tracking (MPPT) algorithms have developed towards multi-task intelligent computing, processors in photovoltaic power generation control systems must be capable of achieving a higher performance. However, the challenges posed by the complex environment of photovoltaic fields with regard to processor reliability cannot be overlooked. To address these issues, we proposed a novel approach. Our approach uses error rate and performance as switching metrics and performs joint statistics to achieve efficient adaptive switching. Based on this, our work designed a redundancy-mode switchable three-core processor system to balance performance and reliability. Additionally, by analyzing the relationship between performance and reliability, we proposed optimization methods to improve reliability while ensuring a high performance was maintained. Finally, we designed an error injection method and verified the system’s reliability by analyzing the error rate probability model in different scenarios. The results of the analysis show that compared with the traditional MPPT controller, the redundancy mode switchable multi-core processor system proposed in this paper exhibits a reliability approximately 5.58 times that of a non-fault-tolerant system. Furthermore, leveraging the feature of module switching, the system’s performance has been enhanced by 26% compared to a highly reliable triple modular redundancy systems, significantly improving the system’s reliability while ensuring a good performance is maintained.

Background Autoimmune necrotizing myopathy with anti-signal recognition particle antibodies (ANM-SRP) is regarded as refractory myositis, whereby some patients respond poorly to conventional immunosuppression and require B cell depletion treatment. This study aimed to evaluate factors associated with refractory ANM-SRP. Results Clinical and pathological data from 48 patients with ANM-SRP were collected. We followed up clinical symptoms and image changes over 12 months. Univariate and multivariate analyses were undertaken to determine the associations between variables of interest and poor response to therapy. Refractory ANM-SRP appeared in 32.5% of patients who showed no or minimal improvement after 12 months of steroid therapy. The clinical risk factors for refractory patients were being male (OR, 19.57; P  < 0.001), severe muscle weakness (OR, 7.51; P  < 0.001) and concurrent interstitial lung disease (OR, 39.70; P  < 0.001). The imaging refractory-related factor was the fatty infiltration rate of thigh muscles over 3 months ( P  = 0.022) and the pathological factor associated with refractory ANM-SRP was the high expression of B cell activating factor receptor (BAFF-R) in muscle ( P  = 0.036). Conclusion Being male, severe muscle weakness, concurrent interstitial lung disease, quick development of muscle fatty infiltration and more BAFF-R and B lymphocyte infiltration in muscle indicate a poor response to immunosuppressive therapy in patients with ANM-SRP.

To better understand the effect of the size of hole defects on the mechanical properties of a rock mass, the two-dimensional particle flow discrete element code (PFC2D) is applied to establish rock mass models with single circular hole defects of different diameters. Uniaxial compressive strength (UCS) tests are conducted on each model by only taking the defect size (area) as a variable. This study analyzes each model’s stress-strain, contact force chain, crack evolution, meso-damage and failure, and mechanical properties. The results showed that with the size enlargement of the circular hole defects, each model’s UCS and elastic modulus gradually decrease, and the defect size is negatively correlated with the mechanical strength of the rock samples. The size of the hole defects affects the entire process of contact force chain and crack evolution. The larger the aperture dimension of the circular hole defects in each model, the greater the concentration degree of the contact force chain, the earlier the crack initiation, and the higher the degree of crack coalescence in the post-peak stage. The number of cracks decreases as the hole size increases, and the model is more prone to failure. Rock models’ strength and failure characteristics with different numbers and arrangements of hole defects are discussed under the same defect area condition.

The largest ultra-deep (>6000 m) strike-slip fault-controlled oilfield in the world is found in the Tarim Basin of Northwestern China. The localized fractured reservoirs are the major production targets along the strike-slip fault zones. Different from its use in the primary porous-type reservoirs, however, the conventional technology is not favorable for use in oil/gas development in Ordovician carbonate reservoirs. For this reason, high-density seismic acquisition and high-resolution seismic processing were carried out to provide high-precision data for fault and fractured reservoir identification. In addition, the multi-filtering process and the maximum likelihood method are typically used to identify small faults and fault segments along a strike-slip fault zone. Further, seismic facies-constrained inversion and amplitude attributes are favorable for large fracture-cave reservoir description. With the advancements in seismic technology, the high and stable production well ratio has been doubled in the “sweet spots” of fractured reservoir optimization, and the first ultra-deep strike-slip fault-controlled oilfield with an annual oil production of over 1 million tons has been realized, achieving economic development in the ultra-deep fractured reservoirs. However, unstable production and high rates of production decline are still significant challenges in the economic exploitation of the ultra-deep fractured reservoirs. Seismic technology requires further improvement for the description of small fractured reservoirs and matrix reservoirs, as well as reservoir connectivity prediction and hydrocarbon detection in the deep subsurface.