Search Results (64)

In recent years, segmentation for medical applications using Magnetic Resonance Imaging (MRI) has received increasing attention. Working in this field has emerged as an ambitious task and a major challenge for researchers; particularly, brain tumor segmentation from MRI is a crucial task for accurate diagnosis, treatment planning, and patient monitoring. With the rapid development of deep learning methods, significant improvements have been made in medical image segmentation. Convolutional Neural Networks (CNNs), such as U-Net, have shown excellent performance in capturing local spatial features. However, these models cannot explicitly capture long-range dependencies. Therefore, Vision Transformers have emerged as an alternative segmentation method recently, as they can exploit long-range correlations through the self-attention mechanism (MSA). Despite their effectiveness, ViTs require large annotated datasets and may compromise fine-grained spatial details. To address these problems, we propose a novel hybrid approach for brain tumor segmentation that combines a 3D U-Net with a 3D Vision Transformer (ViT3D), aiming to jointly exploit local feature extraction and global context modeling. Additionally, we developed an effective fusion method that uses upsampling and convolutional refinement to improve multi-scale feature integration. Unlike traditional fusion approaches, our method explicitly refines spatial details while maintaining global dependencies, improving the quality of tumor border delineation. We evaluated our approach on the BraTS 2020 dataset, achieving a global accuracy score of 99.56%, an average Dice similarity coefficient (DSC) of 77.43% (corresponding to the mean across the three tumor subregions), with individual Dice scores of 84.35% for WT, 80.97% for TC, and 66.97% for ET, and an average Intersection over Union (IoU) of 71.69%. These extensive experimental results demonstrate that our model not only localizes tumors with high accuracy and robustness but also outperforms a selection of current state-of-the-art methods, including U-Net, SwinUnet, M-Unet, and others. Full article

With the rapid growth of urban populations and the expansion of metro networks, accurate energy consumption prediction has become a critical task for optimizing metro operations and supporting low-carbon city development. Traditional statistical and machine learning methods often struggle to model the complex, nonlinear, and time-varying nature of metro energy data. To address these challenges, this paper proposes MTMM, a novel hybrid model that integrates the multi-head attention mechanism of the Transformer with the efficient, state-space-based Mamba architecture. The Transformer effectively captures long-range temporal dependencies, while Mamba enhances inference speed and reduces complexity. Additionally, the model incorporates multivariate energy features, leveraging the correlations among different energy consumption types to improve predictive performance. Experimental results on real-world data from the Guangzhou Metro demonstrate that MTMM significantly outperforms existing methods in terms of both MAE and MSE. The model also shows strong generalization ability across different prediction lengths and time step configurations, offering a promising solution for intelligent energy management in metro systems. Full article

The ecological environment of Inner Mongolia constitutes a critical component of China’s ecological civilization construction. To comprehensively assess and monitor ecological quality dynamics in this region, this study employed MODIS remote sensing data products (2000–2020) and derived four key indicators, —vegetation index (NDVI), wetness index (WET), build-up and soil index (NDBSI), and land surface temperature (LST)—via the Google Earth Engine (GEE) platform. A Remote Sensing-based Ecological Index (RSEI) was constructed using principal component analysis (PCA) to establish an annual long-term time series, thereby eliminating subjective bias from artificial weight assignment. Integrated methodologies—including Theil–Sen Median and Mann–Kendall trend analysis, Hurst exponent, and geographical detector—were applied to investigate the spatiotemporal evolution of ecological quality in Inner Mongolia and its responses to climatic and anthropogenic drivers. This study proposes a novel framework for large-scale ecological quality assessment using remote sensing. Key findings include the following: The mean RSEI value of 0.41 (2000–2020) indicates an overall improving trend in ecological quality. Areas with ecological improvement and degradation accounted for 76.06% and 23.84% of the region, respectively, exhibiting a spatial pattern of “northwestern improvement versus southeastern degradation.” Pronounced regional disparities were observed: optimal ecological conditions prevailed in the Greater Khingan Range (northeast), while the Alxa League (southwest) exhibited the poorest conditions. Northwestern improvement was primarily driven by increased precipitation, rising temperatures, and conservation policies, whereas southeastern degradation correlated with rapid urbanization and intensified socioeconomic activities. Our results demonstrate that MODIS-derived RSEI effectively enables large-scale ecological monitoring, providing a scientific basis for regional green development strategies. Full article

Background/Objectives: We present a novel approach for detecting generalized sleep pathologies through the fractal analysis of single-channel electroencephalographic (EEG) signals. We propose that the fractal scaling exponent of permutation entropy time series serves as a robust biomarker of pathological sleep patterns, capturing alterations in brain dynamics across multiple disorders. Methods: Using two public datasets (Sleep-EDF and CAP Sleep Database) comprising 200 subjects (112 healthy controls and 88 patients with various sleep pathologies), we computed the fractal scaling of the permutation entropy of these signals. Results: The results demonstrate significantly reduced scaling exponents in pathological sleep compared to healthy controls (mean = 1.24 vs. 1.06, $p<0.001$), indicating disrupted long-range temporal correlations in neural activity. The method achieved 90% classification accuracy for rapid-eye-movement (REM) sleep behavior disorder (F1-score: 0.89) and maintained 74% accuracy when aggregating all pathologies (insomnia, narcolepsy, sleep-disordered breathing, etc.). Conclusions: The advantages of this approach, including compatibility with single-channel EEG (enabling potential wearable applications), independence from sleep-stage annotations, and generalizability across recording montages and sampling rates, stablish a framework for non-specific sleep pathology detection. This is a computationally efficient method that could transform screening protocols and enable earlier intervention. The robustness of this biomarker could enable straightforward clinical applications for common sleep pathologies as well as diseases associated with neurodegenerative conditions. Full article

Hydrocortisone is a typical glucocorticoid commonly used in livestock production; however, its overuse can result in hormone residues in milk. Long-term consumption of such milk may lead to a series of health issues. Therefore, the timely and rapid detection of hydrocortisone in milk is crucial for protecting human health. In this study, a magnetic particle-based direct chemiluminescence immunoassay (MP-DCLIA) incorporating a streptavidin–biotin signal amplification system was developed for the rapid and high-throughput detection of hydrocortisone in milk. Automated operations reduce human error and enhance the accuracy and repeatability of tests. The assay can be completed in 12 min with a linear detection range of 13.09–261.71 μg/L, a limit of detection (LOD) of 4.94 μg/L, a limit of quantification (LOQ) of 14.84 μg/L, and intra- and inter-batch variations of less than 5%. The method demonstrated stability and exhibited no cross-reactivity with structural analogues. Spiked recoveries of milk samples ranged from 85.85% to 100.30%, with results strongly correlating with those obtained from LC-MS/MS. The MP-DCLIA offers rapidity, high efficiency, stability, and precision, making it a promising tool for practical testing applications. Full article

Traditional single neural network-based geo-localization methods for cross-view imagery primarily rely on polar coordinate transformations while suffering from limited global correlation modeling capabilities. To address these fundamental challenges of weak feature correlation and poor scene adaptation, we present a novel framework termed ICT-Net (Integrated CNN-Transformer Network) that synergistically combines convolutional neural networks with Transformer architectures. Our approach harnesses the complementary strengths of CNNs in capturing local geometric details and Transformers in establishing long-range dependencies, enabling comprehensive joint perception of both local and global visual patterns. Furthermore, capitalizing on the Transformer’s flexible input processing mechanism, we develop an attention-guided non-uniform cropping strategy that dynamically eliminates redundant image patches with minimal impact on localization accuracy, thereby achieving enhanced computational efficiency. To facilitate practical deployment, we propose a deep embedding clustering algorithm optimized for rapid parsing of geo-localization information. Extensive experiments demonstrate that ICT-Net establishes new state-of-the-art localization accuracy on the CVUSA benchmark, achieving a top-1 recall rate improvement of 8.6% over previous methods. Additional validation on a challenging real-world dataset collected at Beihang University (BUAA) further confirms the framework’s effectiveness and practical applicability in complex urban environments, particularly showing 23% higher robustness to vegetation variations. Full article

Rapid climate change has significantly impacted species distribution patterns, necessitating a comprehensive understanding of dominant tree dynamics for effective forest resource management and utilization. The Camellia subgenus Camellia, a widely distributed taxon in subtropical China, represents an ecologically and economically important group of woody plants valued for both oil production and ornamental purposes. In this study, we employed the BIOMOD2 ensemble modeling framework to investigate the spatial distribution patterns and range dynamics of the subgenus Camellia under projected climate change scenarios. Our analysis incorporated 1455 georeferenced occurrence records from 15 species, following the filtering of duplicate points, along with seven bioclimatic variables selected after highly correlated factors were eliminated. The ensemble model, which integrates six single species distribution models, demonstrated robust predictive performance, with mean true skil l statistic (TSS) and area under curve (AUC) values exceeding 0.8. Our results identified precipitation of the coldest quarter (Bio19) and temperature seasonality (Bio4) as the primary determinants influencing species distribution patterns. The center of species richness for the subgenus Camellia was located in the Nanling Mountains and eastern Guangxi Zhuang Autonomous Region. The projections indicate an overall expansion of suitable habitats for the subgenus under future climate conditions, with notable scenario-dependent variations: distribution hotspots are predicted to increase by 8.86% under the SSP126 scenario but experience a 2.53% reduction under the SSP585 scenario. Furthermore, a westward shift in the distribution centroid is anticipated. To ensure long-term conservation of Camellia genetic resources, we recommend establishing a germplasm conservation center in the Nanling Mountains region, which represents a critical biodiversity hotspot for this taxon. Full article

High-flying insects that exploit tropospheric winds can disperse over far greater distances in a single generation than species restricted to below-canopy flight. However, the ecological consequences of such long-range dispersal remain poorly understood. For example, high-altitude dispersal may facilitate more rapid range shifts in these species and reduce their sensitivity to habitat fragmentation, in contrast to low-flying insects that rely more on terrestrial patch networks. Previous studies have primarily used surface-level variables with limited spatial coverage to explore dispersal timing and movement. In this study, we introduce a novel application of niche modelling to insect aeroecology by examining the relationship between a comprehensive set of atmospheric conditions and high-flying insect activity in the troposphere, as detected by weather surveillance radars (WSRs). We reveal correlations between large-scale dispersal events and atmospheric conditions, identifying key variables that influence dispersal behaviour. By incorporating high-altitude atmospheric conditions into niche models, we achieve significantly higher predictive accuracy compared with models based solely on surface-level conditions. Key predictive factors include the proportion of arable land, altitude, temperature, and relative humidity. Full article

The utilization of recombinant human growth hormone therapy in pediatric populations, originally approved to treat diseases of growth hormone deficiency, has expanded to encompass a broader range of indications, leading to a threefold increase in its utilization in the last two decades. However, concerns regarding its safety, particularly those that are orthopedic in nature, have grown alongside its increasing popularity. Growth hormone usage has been reported to predispose patients to a multitude of common orthopedic conditions, including carpal tunnel syndrome, Legg–Calve–Perthes disease, little league shoulder, Osgood–Schlatter disease, osteochondritis dissecans, scoliosis, Sever’s disease, and slipped femoral capital epiphysis. The pathways by which growth hormone therapy can precipitate orthopedic pathology has been shown to be multifactorial, involving mechanisms such as hormonal changes, growth plate instability, rapid growth, and increased susceptibility to overuse injury. This review examines the orthopedic consequences of growth hormone therapy in pediatric patients by discussing these potential pathophysiologic mechanisms of injury and analyzing subsequent clinical manifestations. By examining processes underlying these complications, we highlight the need for orthopedic surveillance and management in children receiving GHT, particularly those with pre-existing musculoskeletal comorbidities or high levels of physical activity. Our findings underscore the importance of a multidisciplinary approach involving co-management by pediatricians, endocrinologists, and orthopedic surgeons to optimize safety and outcomes for these patients. Directions for future research include correlating pathophysiologic mechanisms to injury patterns, investigating long-term complications in recently approved growth hormone therapy indications, and informing clinical guidelines on the management of orthopedic injuries in this patient population. Full article

The rapid development of sensor technology has made multi-modal remote sensing data valuable for land cover classification due to its diverse and complementary information. Many feature extraction methods for multi-modal data, combining light detection and ranging (LiDAR) and hyperspectral imaging (HSI), have recognized the importance of incorporating multiple spatial scales. However, effectively capturing both long-range global correlations and short-range local features simultaneously on different scales remains a challenge, particularly in large-scale, complex ground scenes. To address this limitation, we propose a multi-scale graph encoder–decoder network (MGEN) for multi-modal data classification. The MGEN adopts a graph model that maintains global sample correlations to fuse multi-scale features, enabling simultaneous extraction of local and global information. The graph encoder maps multi-modal data from different scales to the graph space and completes feature extraction in the graph space. The graph decoder maps the features of multiple scales back to the original data space and completes multi-scale feature fusion and classification. Experimental results on three HSI-LiDAR datasets demonstrate that the proposed MGEN achieves considerable classification accuracies and outperforms state-of-the-art methods. Full article

With the advantages of a high specificity, a long half-life, and a high safety, the use of antibody biologic drugs, including camrelizumab, has been rapidly increasing in clinical practice. Camrelizumab, an immune checkpoint inhibitor and humanized monoclonal antibody, is used to treat several advanced solid cancers. Measuring its concentration supports personalized dosage adjustments, influences treatment decisions for patients, strengthens the control of disease activity through therapeutic drug monitoring, and helps evaluate and prevent drug interactions in combination therapy. Because antibodies are present in complex biological matrices, quantifying monoclonal antibody drugs is challenging, and must rely on precise, selective, and reliable analytical methods. In this study, a quadrupole time-of-flight mass spectrometry TripleTOF 6600+ (AB SCIEX, Framingham, MA, USA) system equipped with a Turbo V ion source was used for the qualitative analysis of monoclonal antibodies using the data-dependent acquisition (IDA) MS/MS mode, followed by quantitative analysis using a targeted MRMHR workflow. This method showed a good linear relationship within the range of 4–160 μg/mL, with a correlation coefficient of R² ≥ 0.996. It demonstrated an acceptable accuracy (88.95–101.18%) and precision (≤15%). Furthermore, the lower limit of quantification was found to be 4 μg/mL, with the lowest detection limit of 0.3217 μg/mL, indicating that this method is rapid, accurate, and reliable for the quantitative analysis of camrelizumab in human serum. Full article

Traditional soil salinity measurement methods are generally complex and labor-intensive, restricting the long-term monitoring of soil salinity, particularly in arid areas. In this context, the soil salt content (SSC) data from farms in the Heihe River Basin in Northwest China were collected in three consecutive years (2021, 2022, and 2023). In addition, the spectral reflectance and texture features of different sampling sites in the study area were extracted from long-term unmanned aerial vehicle (UAV) multispectral images to replace the red and near-infrared bands with a newly introduced red edge band. The spectral index was calculated in this study before using four sensitive variable combinations to predict soil salt contents. A Pearson correlation analysis was performed in this study to screen 57 sensitive features. In addition, 36 modeling scenarios were conducted based on the Extreme Gradient Boosting (XGBoost Implemented using R language 4.3.1), Backpropagation Neural Network (BPNN), and Random Forest (RF) algorithms. The most optimal algorithms for predicting the soil salt contents in farmland located in the Heihe River Basin, in the arid region of Northwest China, were determined. The results showed a higher prediction accuracy for the XGBoost algorithm than the RF and BPNN algorithms, accurately reflecting the actual soil salt contents in the arid area. On the other hand, the most accurate predicted soil salt contents were obtained in 2023 using the XGBoost algorithm, with coefficient of determination (R²), root mean square error (RMSE), and mean absolute error (MAE) ranges of 0.622–0.820, 0.086–0.157, and 0.078–0.134, respectively, whereas the most stable prediction results were obtained using the collected data in 2022. From the perspective of different sensitive variable input combinations, the implementation of the XGBoost algorithm using the spectral index–spectral reflectance–texture feature input combination resulted in comparatively higher prediction accuracies than those of the other variable combinations in 2022 and 2023. Specifically, the R², RMSE, and MAE values obtained using the spectral index–spectral reflectance–texture feature input combination were 0.674, 0.133, and 0.086 in 2022 and 0.820, 0.165, and 0.134 in 2023, respectively. Therefore, our results demonstrated that the spectral index–spectral reflectance–texture feature was the optimal sensitive variable input combination for the machine learning algorithms, of which the XGBoost algorithm is the most optimal model for predicting soil salt contents. The results of this study provide a theoretical basis for the rapid and accurate prediction of soil salinity in arid areas. Full article

Potassium (K) is an essential nutrient for the growth and development of plants. Root hairs are the main parts of plants that absorb K⁺. The regulation of plant root hair growth in response to a wide range of environmental stresses is crucially associated with the dynamics of actin filaments, and the thick actin bundles at the apical and sub-apical regions are essential for terminating the rapid elongation of root hair cells. However, the dynamics and roles of actin filaments in root hair growth in plants’ response to low K⁺ stress are not fully understood. Here, we revealed that root hairs grow faster and longer under low K⁺ stress than the control conditions. Compared to control conditions, the actin filaments in the sub-apex of fast-growing wild-type root hairs were longer and more parallel under low K⁺ stress, which correlates with an increased root hair growth rate under low K⁺ stress; the finer actin filaments in the sub-apex of the early fully grown Col-0 root hairs under low K⁺ stress, which is associated with low K⁺ stress-induced root hair growth time. Further, Arabidopsis thaliana actin bundling protein Villin1 (VLN1) and Villin4 (VLN4) was inhibited and induced under low K⁺ stress, respectively. Low K⁺ stress-inhibited VLN1 led to decreased bundling rate and thick bundle formation in the early fully grown phase. Low K⁺ stress-induced VLN4 functioned in keeping long filaments in the fast-growing phase. Furthermore, the analysis of genetics pointed out the involvement of VLN1 and VLN4 in the growth of root hairs under the stress of low potassium levels in plants. Our results provide a basis for the dynamics of actin filaments and their molecular regulation mechanisms in root hair growth in response to low K⁺ stress. Full article

The economic belt on the north slope of the Tianshan Mountains is a highly productive area in Xinjiang, but with the rapid development of the economy and industry and the acceleration of urbanization in recent years, the fragile ecological environment in the region has further deteriorated. Exploring shifts in land utilization across different eras and regions, along with the transformation of terrain configurations, provides key perspectives that can propel sustainable societal and environmental growth within this particular area. The research analyzed four periods (1990, 2000, 2010, 2020) of remote sensing image data combined with field monitoring data using methods such as land use variability, landscape pattern index, and grey relational model. Focusing on investigating the dynamics of the ecological environment in high-intensity human activity areas, examining alterations in land use patterns over time and space, transitions in land use types, and trends in landscape pattern indices. (1) The dominant land environments situated in the economic zone adjacent to the northern base of the Tianshan mountain range encompass extensive expanses of grassy plains and unexploited landscapes, making up 45% and 38% of the area, correspondingly. The single dynamic change degree of construction land was the largest due to the implementation of long-term land development and urbanization policies. Land use transfer change mainly occurred among cultivated land, grassland, forestland, and unused land. With strong human activities, the construction land area has expanded by 145.16% (2089.7 km²), and this number is still increasing. (2) The spatial landscape structure on the north slope of Tianshan Mountain is becoming more complicated and diversified; the cities with the highest degree of fragmentation were concentrated in the middle and western sections. Grassland is the most dominant patch type in the landscape. The shape of patches tends to be irregular and complex in general, and the fragmentation degree and dispersion degree of landscape patches are enhanced as the proportion of different landscape types increases. (3) Grey correlation analysis indicates that grasslands, cultivated land, and unused land are key elements in the landscape pattern changes on the northern slope of the Tianshan Mountains. Central urban agglomeration is an area with strong landscape pattern changes, and ecological protection should be emphasized while promoting economic development. Full article

Seasonal variations in glycaemic patterns in children and young adults affected by type 1 diabetes are currently poorly studied. However, the spread of Flash Glucose Monitoring (FGM) and continuous glucose monitoring (CGM) systems and of dedicated platforms for the synchronization and conservation of CGM reports allows an efficient approach to the comprehension of these phenomena. Moreover, the impact that environmental parameters may have on glycaemic control takes on clinical relevance, implying a need to properly educate patients and their families. In this context, it can be investigated how blood glucose patterns in diabetic patients may have a link to outdoor temperatures. Therefore, in this study, the relationship between outdoor temperatures and glucose levels in diabetic patients, aged between 4 and 21 years old, has been analysed. For a one-year period (Autumn 2022–Summer 2023), seasonal variations in their CGM metrics (i.e., time in range (TIR), Time Above Range (TAR), Time Below Range (TBR), and coefficient of variation (CV)) were analysed with respect to atmospheric temperature. The results highlight a negative correlation between glucose in diabetic patients and temperature patterns (R value computed considering data for the entire year; Ry = −0.49), behaviour which is strongly confirmed by the analysis focused on the July 2023 heatwave (R = −0.67), which shows that during heatwave events, the anticorrelation is accentuated. The diurnal analysis shows how glucose levels fluctuate throughout the day, potentially correlating with atmospheric diurnal temperature changes in addition to the standard trend. Data captured during the July 2023 heatwave (17–21 July 2023) highlight pronounced deviations from the long-term average, signalling the rapid effects of extreme temperatures on glucose regulation. Our findings underscore the need to integrate meteorological parameters into diabetes management and clinical trial designs. These results suggest that structured diabetes self-management education of patients and their families should include adequate warnings about the effects of atmospheric temperature variations on the risk of hypoglycaemia and about the negative effects of excessive therapeutic inertia in the adjustment of insulin doses. Full article