Explore the vast range of titles available.

Coastal groundwater is an important resource in the developed region associated with human health and sustainable economic development. To identify the origins of salinity and evaluate the impact of water-rock interactions, seawater intrusion (SWI), and evaporation on groundwater in the coastal areas of Zhejiang and Fujian provinces, a comprehensive investigation was performed. Meanwhile, nitrate and fluoride indicators resulting from the anthropogenic activity and SWI were also considered. At last, the water quality index (WQI) of coastal groundwater was evaluated with geochemical and multivariate statistical methods. The results indicated that (1) the groundwater in coastal areas of Zhejiang and Fujian provinces has been affected by SWI to varying degrees. The analysis of selected ion ratios (Na+/Cl- and Br-/Cl-) and isotopic compositions showed that SWI is the predominant cause of increasing salinity in the groundwater of Zhejiang Province, while the cause is water-rock interactions (ion exchange and mineral weathering) in Fujian Province. The hydrochemical evolution path of groundwater in Zhejiang Province is Ca/Mg-HCO3 to Na-Cl, while a different pattern of Ca/Mg-HCO3 to Na (Mg/Ca)-Cl occurs in Fujian Province. However, the trend of SWI development in both provinces was freshening. (2) Nitrification, sewage infiltration, and SWI increased the NO3- content in groundwater. Some of the NO3- concentration in Fujian Province exceeds the standard, and the nitrogen pollution was more serious than in Zhejiang Province. The F- content in coastal groundwater was affected by SWI and mineral dissolution; the F- content in Zhejiang Province was higher than in Fujian Province, which was close to the groundwater standard limit. The average WQI value of Zhejiang was 103.61, and the WQI of Fujian was 61.69, indicating that the coastal groundwater quality in Fujian Province was better than in Zhejiang Province. The results of the study revealed the impact of SWI and anthropogenic activity on groundwater in the southern coastal zone of China and will be valuable for sustainable groundwater resource management.

In large public places such as railway stations and airports, dense pedestrian detection is important for safety and security. Deep learning methods provide relatively effective solutions but still face problems such as feature extraction difficulties, image multi-scale variations, and high leakage detection rates, which bring great challenges to the research in this field. In this paper, we propose an improved dense pedestrian detection algorithm GR-yolo based on Yolov8. GR-yolo introduces the repc3 module to optimize the backbone network, which enhances the ability of feature extraction, adopts the aggregation–distribution mechanism to reconstruct the yolov8 neck structure, fuses multi-level information, achieves a more efficient exchange of information, and enhances the detection ability of the model. Meanwhile, the Giou loss calculation is used to help GR-yolo converge better, improve the detection accuracy of the target position, and reduce missed detection. Experiments show that GR-yolo has improved detection performance over yolov8, with a 3.1% improvement in detection means accuracy on the wider people dataset, 7.2% on the crowd human dataset, and 11.7% on the people detection images dataset. Therefore, the proposed GR-yolo algorithm is suitable for dense, multi-scale, and scene-variable pedestrian detection, and the improvement also provides a new idea to solve dense pedestrian detection in real scenes.

Neonatal herpes simplex virus type 1 (HSV-1) infections contribute to various neurodevelopmental disabilities and the subsequent long-term neurological sequelae into the adulthood. However, further understanding of fetal brain development and the potential neuropathological effects of the HSV-1 infection are hampered by the limitations of existing neurodevelopmental models due to the dramatic differences between humans and other mammalians. Here we generated in vitro neurodevelopmental disorder models including human induced pluripotent stem cell (hiPSC)-based monolayer neuronal differentiation, three-dimensional (3D) neuroepithelial bud, and 3D cerebral organoid to study fetal brain development and the potential neuropathological effects induced by the HSV-1 infections. Our results revealed that the HSV-1-infected neural stem cells (NSCs) exhibited impaired neural differentiation. HSV-1 infection led to dysregulated neurogenesis in the fetal neurodevelopment. The HSV-1-infected brain organoids modelled the pathological features of the neurodevelopmental disorders in the human fetal brain, including the impaired neuronal differentiation, and the dysregulated cortical layer and brain regionalization. Furthermore, the 3D cerebral organoid model showed that HSV-1 infection promoted the abnormal microglial activation, accompanied by the induction of inflammatory factors, such as TNF-α, IL-6, IL-10, and IL-4. Overall, our in vitro neurodevelopmental disorder models reconstituted the neuropathological features associated with HSV-1 infection in human fetal brain development, providing the causal relationships that link HSV biology with the neurodevelopmental disorder pathogen hypothesis.

Liver cancer is one of the malignancies with high mortality rates worldwide, and its timely detection and accurate diagnosis are crucial for improving patient prognosis. To address the limitations of traditional image segmentation techniques and the U-Net network in capturing fine image features, this study proposes an improved model based on the U-Net architecture, named RHEU-Net. By replacing traditional convolution modules in the encoder and decoder with improved residual modules, the network’s feature extraction capabilities and gradient stability are enhanced. A Hybrid Gated Attention (HGA) module is integrated before the skip connections, enabling the parallel processing of channel and spatial attentions, optimizing the feature fusion strategy, and effectively replenishing image details. A Multi-Scale Feature Enhancement (MSFE) layer is introduced at the bottleneck, utilizing multi-scale feature extraction technology to further enhance the expression of receptive fields and contextual information, improving the overall feature representation effect. Testing on the LiTS2017 dataset demonstrated that RHEU-Net achieved Dice scores of 95.72% for liver segmentation and 70.19% for tumor segmentation. These results validate the effectiveness of RHEU-Net and underscore its potential for clinical application.

The increase in the size of the city and the increase in population mobility have greatly increased the number of vehicles on the road, and at the same time brought considerable challenges to the traffic management department. In recent years, more and more experts and scholars have devoted themselves to applying sensors, network communication and dynamic adaptive technologies to road traffic management systems. At present, people have not completely overcome all kinds of complex problems in traffic supervision. The complexity of traffic information and the defects of identification algorithms have brought great challenges to intelligent traffic management. This article has launched a research on the intelligent Internet of Things traffic management system, with the detection and recognition of stationary vehicles and seat belts as the key analysis targets. When monitoring stationary vehicles, this paper replaces the background difference algorithm commonly used in dynamic vehicle detection with a new recognition algorithm. From the experimental results, the average detection accuracy of the new algorithm is 96.77% higher than the previous 87.56%. When studying the driver's seat belt detection, this paper combines the YOLOv3 target detection algorithm and the lightweight network structure, and proposes a driver-oriented positioning algorithm. With the increase in the number of lightweight templates, the accuracy of the positioning algorithm has increased from 80.57 to 99.98%. But on the other hand, the detection speed has also changed from 78 to 69 frames/s.

Background Genetic variability and liability to develop drug-resistant mutations are the main characteristics of HIV-1, which can not only increase the risk of antiretroviral treatment (ART) failure, but also can lead to the spread of resistant strains. We aim to investigate the distribution of HIV-1 genotypes and prevalence of pretreatment drug resistance (PDR) in ART-naïve HIV-1 infected patients in Shanghai China. Methods A cross-sectional study was performed among the newly diagnosed ART-naive HIV-1 infected patients during the period from January 2017 to November 2017 in Shanghai Public Health Clinical Center. The target fragment of 1316 bp in the pol gene spanning the reverse transcriptase and protease regions was amplified using a nested polymerase chain reaction. HIV-1 genotypes were determined by phylogenetic analysis, and PDR associated mutations were determined according to Stanford University HIV Drug Resistance Database ( http://hivdb.stanford.edu/ ). Results We successfully amplified pol gene sequences from blood samples of 317 patients, of whom 95.3% were male, and 68.8% were men who have sex with men. The median age was 33 years; and the median CD4 count was 275 cells/μL. The predominant HIV-1 genotype was circulating recombinant form (CRF) 01_AE (53.0%, 168/317), followed by CRF07_BC (29.7%, 94/317), B (7.6%, 24/317), CRF08_BC (1.9%, 6/317), CRF55_01B (1.9%, 6/317), CRF 59_01B (0.9%, 3/317). In addition, 5% (16/317) HIV-1 strains were identified as other subtypes or CRFs/URFs (unique recombinant forms). The overall prevalence of PDR was 17.4% (55/317). PDR frequency to non-nucleoside reverse transcriptase inhibitor (NNRTI, 16.4%) was much higher than that to nucleoside reverse transcriptase inhibitor (NRTI, 4.7%) and protease inhibitor (PI, 0.6%). The most common HIV-1 mutation pattern for NNRTI and NRTI were V179D/E (10.1%, 32/317) and M184 V (2.8%, 9/317), respectively. About half (49.1%, 27/55) of the HIV-1 strains with mutation presented as potential low-level resistant to NNRTI attributed to V179D/E. Conclusion The distribution of HIV-1 genotypes in Shanghai China is diverse and complex. The high prevalence of PDR highlights the significance of baseline HIV-1 drug resistance testing. Non-NNRTI-containing regimen may be the preferred initial therapy for newly diagnosed HIV-1 patients in Shanghai in the absence of PDR test results.

Scene text detection is an important research field in computer vision, playing a crucial role in various application scenarios. However, existing scene text detection methods often fail to achieve satisfactory results when faced with text instances of different sizes, shapes, and complex backgrounds. To address the challenge of detecting diverse texts in natural scenes, this paper proposes a multi-scale natural scene text detection method based on attention feature extraction and cascaded feature fusion. This method combines global and local attention through an improved attention feature fusion module (DSAF) to capture text features of different scales, enhancing the network’s perception of text regions and improving its feature extraction capabilities. Simultaneously, an improved cascaded feature fusion module (PFFM) is used to fully integrate the extracted feature maps, expanding the receptive field of features and enriching the expressive ability of the feature maps. Finally, to address the cascaded feature maps, a lightweight subspace attention module (SAM) is introduced to partition the concatenated feature maps into several sub-space feature maps, facilitating spatial information interaction among features of different scales. In this paper, comparative experiments are conducted on the ICDAR2015, Total-Text, and MSRA-TD500 datasets, and comparisons are made with some existing scene text detection methods. The results show that the proposed method achieves good performance in terms of accuracy, recall, and F-score, thus verifying its effectiveness and practicality.

Convolutional neural networks (CNNs) have made significant progress in the field of facial expression recognition (FER). However, due to challenges such as occlusion, lighting variations, and changes in head pose, facial expression recognition in real-world environments remains highly challenging. At the same time, methods solely based on CNN heavily rely on local spatial features, lack global information, and struggle to balance the relationship between computational complexity and recognition accuracy. Consequently, the CNN-based models still fall short in their ability to address FER adequately. To address these issues, we propose a lightweight facial expression recognition method based on a hybrid vision transformer. This method captures multi-scale facial features through an improved attention module, achieving richer feature integration, enhancing the network’s perception of key facial expression regions, and improving feature extraction capabilities. Additionally, to further enhance the model’s performance, we have designed the patch dropping (PD) module. This module aims to emulate the attention allocation mechanism of the human visual system for local features, guiding the network to focus on the most discriminative features, reducing the influence of irrelevant features, and intuitively lowering computational costs. Extensive experiments demonstrate that our approach significantly outperforms other methods, achieving an accuracy of 86.51% on RAF-DB and nearly 70% on FER2013, with a model size of only 3.64 MB. These results demonstrate that our method provides a new perspective for the field of facial expression recognition.

Studying the hydrogeochemical processes that control groundwater quality in different aquifers helps to identify the sources of inrush water into coal mines. In this paper, 65 mine water samples were collected from Xiqu coal mine during 2016 and 2017 to determine the hydrogeochemical processes and controlling factors of mine water using hydrochemical methods and multivariate statistical analysis. The mine water is acidic to alkaline in nature. The dominant anions in the mine water are SO42− and HCO3−, and cations were dominated by Ca2+ Na+, and Mg2+. The dominant hydrochemical facies of groundwater were SO4–Na·Ca and SO4–Ca·Mg water types. Based on multivariate statistical analysis, four principal components and two clusters were obtained. Concentrations of SO42− showed significant positive correlations with Ca2+, Mg2+, Fe2+ and Fe3+, and strong negative correlations with HCO3− and pH. The mine water chemistry was dominated by the oxidation of iron-bearing sulfide minerals, cation exchange and the dissolution of silicate, carbonate and sulfate minerals. In addition, human activities such as agriculture are also likely to have influences groundwater chemistry at the mine site.

The world is experiencing an increase in the frequency and intensity of extreme weather events, yet the influences of remote inland extreme weather events on the coastal ecosystem thousands of kilometers away remain poorly understood. Here we tracked the chain ecological effects of an extreme rainfall event in North China from terrestrial rivers to coastal aquaculture area of the eastern Shandong Peninsula. Our data suggest the autumn flood resulted from extreme rainfall event leads to abnormally low turbidity in the North Shandong Coastal Currents and coastal red tide blooms by introducing anomalous freshwater with an exceptionally high nitrogen-to-phosphorus ratio into the Bohai Sea. Lower salinity, stronger light conditions caused by limpid coastal currents, and phosphorus limitation resulting from red tide blooms account for huge kelp loss offshore of the eastern Shandong Peninsula. This study underscores the importance of considering multidisciplinary observation for risk management of unexpected extreme weather events.