Search Results (29)

Segmentation of skin lesions in dermoscopic images is critical for the accurate diagnosis of skin cancers, particularly malignant melanoma, yet it is hindered by irregular lesion shapes, blurred boundaries, low contrast, and artifacts, such as hair interference. Conventional deep learning methods, typically based on UNet or Transformer architectures, often face limitations in regard to fully exploiting lesion features and incur high computational costs, compromising precise lesion delineation. To overcome these challenges, we propose SGNet, a structure-guided network, integrating a hybrid CNN–Mamba framework for robust skin lesion segmentation. The SGNet employs the Visual Mamba (VMamba) encoder to efficiently extract multi-scale features, followed by the Dual-Domain Boundary Enhancer (DDBE), which refines boundary representations and suppresses noise through spatial and frequency-domain processing. The Semantic-Texture Fusion Unit (STFU) adaptively integrates low-level texture with high-level semantic features, while the Structure-Aware Guidance Module (SAGM) generates coarse segmentation maps to provide global structural guidance. The Guided Multi-Scale Refiner (GMSR) further optimizes boundary details through a multi-scale semantic attention mechanism. Comprehensive experiments based on the ISIC2017, ISIC2018, and PH2 datasets demonstrate SGNet’s superior performance, with average improvements of 3.30% in terms of the mean Intersection over Union (mIoU) value and 1.77% in regard to the Dice Similarity Coefficient (DSC) compared to state-of-the-art methods. Ablation studies confirm the effectiveness of each component, highlighting SGNet’s exceptional accuracy and robust generalization for computer-aided dermatological diagnosis. Full article

In order to better understand the micropore structure of shale reservoir in Jiyang Depression, permeability damage test, low temperature nitrogen adsorption and scanning electron microscopy (SEM) were carried out on six cores in the target block. The adsorption isotherms were analyzed by Frenkel–Halsey–Hill (FHH) model, and the fractal dimensions of different layers were calculated. The results show that the shale pore system is mainly composed of organic nanopores, inorganic nanopores and micro-fractures. The inorganic pores are mainly distributed around or inside the mineral particles, while microcracks are commonly found between mineral particles or at the organic–mineral interface. Organic pores are located within or between organic particles. The results of nitrogen adsorption show that the shale pores are mainly H2/H3 hysteresis loops with wedge, plate or ink bottle shapes. The pore structure is highly complex, and the fractal dimension is high. The mean D1 fractal dimension, which represents pore surface roughness, is 2.3788, and the mean D2 fractal dimension, which represents pore structure complexity, is 2.7189. The fractal dimension is positively correlated with specific surface area and total pore volume and negatively correlated with average pore radius. The permeability damage rates of the N layer, B layer, and F layer are 17.39%, 20.2%, and 21.6%, respectively. The contact Angle of the core decreases with the increase in water skiing time. In this study, the micropore structure of different formations in Jiyang Depression is compared and analyzed, which provides valuable insights for the optimization and differentiated development of shale oil and gas resources. Full article

White blood cell (WBC) classification plays a crucial role in hematopathology and clinical diagnostics. However, traditional methods are constrained by limited receptive fields and insufficient utilization of contextual information, which hinders classification performance. To address these limitations, this paper proposes an enhanced WBC classification algorithm, CCE-YOLOv7, which is built upon the YOLOv7 framework. The proposed method introduces four key innovations to enhance detection accuracy and model efficiency: (1) A novel Conv2Former (Convolutional Transformer) backbone was designed to combine the local pattern extraction capability of convolutional neural networks (CNNs) with the global contextual reasoning of transformers, thereby improving the expressiveness of feature representation. (2) The CARAFE (Content-Aware ReAssembly of Features) upsampling operator was adopted to replace conventional interpolation methods, thereby enhancing the spatial resolution and semantic richness of feature maps. (3) An Efficient Multi-scale Attention (EMA) module was introduced to refine multi-scale feature fusion, enabling the model to better focus on spatially relevant features critical for WBC classification. (4) Soft-NMS (Soft Non-Maximum Suppression) was used instead of traditional NMS to better preserve true positives in densely packed or overlapping cell scenarios, thereby reducing false positives and false negatives. Experimental validation was conducted on a WBC image dataset acquired using the Fourier ptychographic microscopy (FPM) system. The proposed CCE-YOLOv7 achieved a detection accuracy of 89.3%, showing a 7.8% improvement over the baseline YOLOv7. Furthermore, CCE-YOLOv7 reduced the number of parameters by 2 million and lowered computational complexity by 5.7 GFLOPs, offering an efficient and lightweight model suitable for real-time clinical applications. To further evaluate model effectiveness, comparative experiments were conducted with YOLOv8 and YOLOv11. CCE-YOLOv7 achieved a 4.1% higher detection accuracy than YOLOv8 while reducing computational cost by 2.4 GFLOPs. Compared with the more advanced YOLOv11, CCE-YOLOv7 maintained competitive accuracy (only 0.6% lower) while using significantly fewer parameters and 4.3 GFLOPs less in computation, highlighting its superior trade-off between accuracy and efficiency. These results demonstrate that CCE-YOLOv7 provides a robust, accurate, and computationally efficient solution for automated WBC classification, with significant clinical applicability. Full article

Background/Objectives: The morphological characteristics of breast tumors play a crucial role in the preliminary diagnosis of breast cancer. However, malignant tumors often exhibit rough, irregular edges and unclear, boundaries in ultrasound images. Additionally, variations in tumor size, location, and shape further complicate the accurate segmentation of breast tumors from ultrasound images. Methods: For these difficulties, this paper introduces a breast ultrasound tumor segmentation network comprising a multi-scale feature acquisition (MFA) module and a multi-input edge supplement (MES) module. The MFA module effectively incorporates dilated convolutions of various sizes in a serial-parallel fashion to capture tumor features at diverse scales. Then, the MES module is employed to enhance the output of each decoder layer by supplementing edge information. This process aims to improve the overall integrity of tumor boundaries, contributing to more refined segmentation results. Results: The mean Dice (mDice), Pixel Accuracy (PA), Intersection over Union (IoU), Recall, and Hausdorff Distance (HD) of this method for the publicly available breast ultrasound image (BUSI) dataset were 79.43%, 96.84%, 83.00%, 87.17%, and 19.71 mm, respectively, and for the dataset of Fujian Cancer Hospital, 90.45%, 97.55%, 90.08%, 93.72%, and 11.02 mm, respectively. In the BUSI dataset, compared to the original UNet, the Dice for malignant tumors increased by 14.59%, and the HD decreased by 17.13 mm. Conclusions: Our method is capable of accurately segmenting breast tumor ultrasound images, which provides very valuable edge information for subsequent diagnosis of breast cancer. The experimental results show that our method has made substantial progress in improving accuracy. Full article

Methicillin-resistant Staphylococcus aureus (MRSA) can spread from animals to humans, but how it adapts to infect both is not fully understood. Our review aimed to determine the prevalence of MRSA in livestock, poultry, and companion animals in different countries, including Bangladesh, the Hong Kong SAR, Mainland China, and Sri Lanka. Articles were collected using PubMed, Embase, Web of Science, Scopus, CINAHL, and Google Scholar. Only prevalence studies that followed the PICO guidelines were included. A random-effects model meta-analysis was used to pool the data. The quality of the evidence and bias were assessed using the GRADEpro and Cochrane collaboration tools. Out of 1438 articles, 69 studies were eligible for meta-analysis. The studies showed significant heterogeneity (I² = 97.00%, p < 0.0001) in the prevalence of MRSA colonization. Therefore, a random-effects model was used to determine the pooled prevalence of MRSA colonization, which was found to be 4.92% (95% CI: 3.79% to 6.18%). Begg’s test (p = 0.0002) and Egger’s test (p = 0.0044) revealed publication bias. Subgroup analysis of the pooled prevalence of MRSA showed a significant difference (p < 0.00001) when the subgroups were divided by country, MRSA detection method, whether pre-enrichment was performed or not, study period, sample collection location, and study population. Although significant factors can partially explain the heterogeneity, it is crucial to recognize the heterogeneity within different subgroups. The pooled prevalence of MRSA was found to vary significantly (p < 0.00001) among the study periods and has increased since the study period of 2020. Therefore, it is crucial to continuously monitor and implement measures to control the spread of MRSA in animals to minimize the risk of transmission to humans. Full article

Background/Objectives: Finotonlimab (SCTI10A) is a humanized anti-PD-1 antibody tested in Phase III trials for several solid tumor types. Methods: This study characterized the in vitro and in vivo efficacy, Fc-mediated effector function, and non-clinical PK/PD properties of finotonlimab. Results: The results demonstrated that finotonlimab is effective in stimulating human T cell function in vitro and exhibits marked antitumor efficacy in vivo using both PD-1-humanized and PBMC-reconstructed mouse models. Additionally, finotonlimab exhibited minimal impact on the activation of effector cells via Fc receptor-dependent pathways, potentially facilitating PD-1⁺ T cell killing. In cynomolgus monkeys, finotonlimab exhibited a nonlinear pharmacokinetic (PK) profile in a dose-dependent manner, and a receptor occupancy rate of approximately 90% was observed at 168 h following a single administration of 1 mg/kg. Finotonlimab’s PK profile (especially C_max) was better than that of marketed antibodies. Following a 13-week successive administration of finotonlimab, a pharmacodynamic analysis revealed that a sustained mean receptor occupancy of PD-1 molecules on circulating T cells remained at or above 93% for up to 8 weeks, even at a dose of 3 mg/kg, and that there were higher antibody accumulations in different dose groups. Conclusions: Taken together, the preclinical findings are promising and provide the groundwork for evaluating the efficacy and pharmacodynamic characteristics of finotonlimab in clinical trials. Full article

Autonomous Fourier Ptychographic Microscopy (FPM) is a technology widely used in the field of pathology. It is compatible with high resolution and large field-of-view imaging and can observe more image details. Red blood cells play an indispensable role in assessing the oxygen-carrying capacity of the human body and in screening for clinical diagnosis and treatment needs. In this paper, the blood cell data set is constructed based on the FPM system experimental platform. Before training, four enhancement strategies are adopted for the blood cell image data to improve the generalization and robustness of the model. A blood cell detection algorithm based on SCD-YOLOv7 is proposed. Firstly, the C-MP (Convolutional Max Pooling) module and DELAN (Deep Efficient Learning Automotive Network) module are used in the feature extraction network to optimize the feature extraction process and improve the extraction ability of overlapping cell features by considering the characteristics of channels and spatial dimensions. Secondly, through the Sim-Head detection head, the global information of the deep feature map (mean average precision) and the local details of the shallow feature map are fully utilized to improve the performance of the algorithm for small target detection. MAP is a comprehensive indicator for evaluating the performance of object detection algorithms, which measures the accuracy and robustness of a model by calculating the average precision (AP) under different categories or thresholds. Finally, the Focal-EIoU (Focal Extended Intersection over Union) loss function is introduced, which not only improves the convergence speed of the model but also significantly improves the accuracy of blood cell detection. Through quantitative and qualitative analysis of ablation experiments and comparative experimental results, the detection accuracy of the SCD-YOLOv7 algorithm on the blood cell data set reached 92.4%, increased by 7.2%, and the calculation amount was reduced by 14.6 G. Full article

Alkaline stress is one of the major abiotic constraints that limits plant growth and development. However, the genetic basis underlying alkaline tolerance in soybean [Glycine max (L.) Merr.] remains largely unexplored. In this study, an integrated genomic analysis approach was employed to elucidate the genetic architecture of alkaline tolerance in a diverse panel of 326 soybean cultivars. Through association mapping, we detected 28 single nucleotide polymorphisms (SNPs) significantly associated with alkaline tolerance. By examining the genomic distances around these significant SNPs, five genomic regions were characterized as stable quantitative trait loci (QTLs), which were designated as qAT1, qAT4, qAT14, qAT18, and qAT20. These QTLs are reported here for the first time in soybean. Seventeen putative candidate genes were identified within the physical intervals of these QTLs. Haplotype analysis indicated that four of these candidate genes exhibited significant allele variation associated with alkaline tolerance-related traits, and the haplotype alleles for these four genes varied in number from two to four. The findings of this study may have important implications for soybean breeding programs aimed at enhancing alkaline tolerance. Full article

Autonomous vehicles operating in public transportation spaces must rapidly and accurately detect all potential hazards in their surroundings to execute appropriate actions such as yielding, lane changing, and overtaking. This capability is a prerequisite for achieving advanced autonomous driving. In autonomous driving scenarios, distant objects are often small, which increases the risk of detection failures. To address this challenge, the MST-YOLOv8 model, which incorporates the C2f-MLCA structure and the ST-P2Neck structure to enhance the model’s ability to detect small objects, is proposed. This paper introduces mixed local channel attention (MLCA) into the C2f structure, enabling the model to pay more attention to the region of small objects. A P2 detection layer is added to the neck part of the YOLOv8 model, and scale sequence feature fusion (SSFF) and triple feature encoding (TFE) modules are introduced to assist the model in better localizing small objects. Compared with the original YOLOv8 model, MST-YOLOv8 demonstrates a 3.43% improvement in precision (P), an 8.15% improvement in recall (R), an 8.42% increase in mAP_0.5, a reduction in missed detection rate by 18.47%, a 70.97% improvement in small object detection AP, and a 68.92% improvement in AR. Full article

Given the diversity of medical images, traditional image segmentation models face the issue of domain shift. Unsupervised domain adaptation (UDA) methods have emerged as a pivotal strategy for cross modality analysis. These methods typically utilize generative adversarial networks (GANs) for both image-level and feature-level domain adaptation through the transformation and reconstruction of images, assuming the features between domains are well-aligned. However, this assumption falters with significant gaps between different medical image modalities, such as MRI and CT. These gaps hinder the effective training of segmentation networks with cross-modality images and can lead to misleading training guidance and instability. To address these challenges, this paper introduces a novel approach comprising a cross-modality feature alignment sub-network and a cross pseudo supervised dual-stream segmentation sub-network. These components work together to bridge domain discrepancies more effectively and ensure a stable training environment. The feature alignment sub-network is designed for the bidirectional alignment of features between the source and target domains, incorporating a self-attention module to aid in learning structurally consistent and relevant information. The segmentation sub-network leverages an enhanced cross-pseudo-supervised loss to harmonize the output of the two segmentation networks, assessing pseudo-distances between domains to improve the pseudo-label quality and thus enhancing the overall learning efficiency of the framework. This method’s success is demonstrated by notable advancements in segmentation precision across target domains for abdomen and brain tasks. Full article

Deep carbonate gas reservoirs exhibit diverse reservoir types and complex seepage patterns. To study the gas production capabilities of different composite reservoir types, we classified the reservoirs of the fourth member of the Dengying Formation in the Anyue Gas Field into high-quality reservoirs (HRs) and poor-quality reservoirs (PRs) based on high-pressure mercury injection (HPMI) experiment results. By varying the differential pressure of the depletion experiment and the connection method, as well as the permeability and water saturation of the composite core, the effects of well location deployment, permeability ratio of the high-quality reservoir and poor-quality reservoir (PRHPR), gas well production pressure difference (GWPPD), and water saturation on the depletion gas production characteristics of the composite reservoir were studied. The research results show that (1) deploying wells on HR enables high gas production rates and ultimate recovery rates; (2) only when the PRHPR falls within a reasonable range (21.88–43.19) can the “dynamic recharge” capability of PR and the high permeability of HR be coordinated to achieve high gas recovery rates; (3) a GWPPD of 3 MPa is optimal, resulting in fast gas production rates and high ultimate recovery rates for PR; (4) high water saturation (≥50%) leads to premature water breakthrough at the well bottom, decreased gas production rate, and sealing of HR and PR reserves by formation water. Combining experimental results with field production data is our next research focus. Our future research focus will be on integrating experimental results with field production data to provide solid theoretical support for the efficient development of this type of gas reservoir. Full article

Pectin is widely used in the food and pharmaceutical industries. However, data on sweet potato pectin extraction and structural property analyses are lacking. Here, for the high-value utilization of agricultural processing waste, sweet potato residue, a byproduct of sweet potato starch processing, was used as raw material. Ammonium oxalate, trisodium citrate, disodium hydrogen phosphate, hydrochloric acid and citric acid were used as extractants for the pectin constituents, among which ammonium oxalate had a high extraction rate of sweet potato pectin, low ash content and high molecular weight. Structural and gelation analyses were conducted on ammonium oxalate-extracted purified sweet potato pectin (AMOP). Analyses showed that AMOP is a rhamnogalacturonan-I-type pectin, with a molecular weight of 192.5 kg/mol. Chemical titration and infrared spectroscopy analysis confirmed that AMOP is a low-ester pectin, and scanning electron and atomic force microscopy demonstrated its linear molecular structure. Gelation studies have revealed that Ca²⁺ is the key factor for gel formation, and that sucrose significantly enhanced gel hardness. The highest AMOP gel hardness was observed at pH 4, with a Ca²⁺ concentration of 30 mg/g, pectin concentration of 2%, and sucrose concentration of 40%, reaching 128.87 g. These results provide a foundation for sweet potato pectin production and applications. Full article

Background and Aims: Type 2 diabetes (T2D) is a global and complex public health challenge, and dietary management is acknowledged as critical in its prevention. Recent studies have highlighted the involvement of micronutrients in T2D pathophysiology; our study aims to assess the association between B vitamin intake and T2D risks and the mediating role of inflammation. Methods: In a prospective cohort design, data on B vitamins intake, including thiamine (B1), riboflavin (B2), niacin (B3), pyridoxine (B6), folate (B9), and cobalamin (B12), was obtained using a validated food frequency questionnaire (FFQ), and blood inflammatory biomarkers were analyzed according to standard protocol in the local hospitals at baseline from 44,960 adults in the Shanghai Suburban Adult Cohort and Biobank (SSACB). Incident T2D cases were identified according to a physician’s diagnosis or medication records from the electronic medical information system. We employed logistic and weighted quantile sum regression models to explore the associations of single and combined levels of B vitamins with T2D and mediation analyses to investigate the effects of inflammation. Results: Negative correlations between B vitamins and T2D were observed in the single-exposure models, except for B3. The analyses of joint exposure (B1, B2, B6, B9, and B12) also showed an inverse association (OR 0.80, 95% CI 0.71 to 0.88), with vitamin B6 accounting for 45.58% of the effects. Further mediation analysis indicated a mediating inflammatory impact, accounting for 6.72% of the relationship. Conclusions: Dietary intake of B vitamins (B1, B2, B6, B9, B12) was associated with a reduced T2D risk partially mediated by inflammation in Shanghai residents. Full article

During the temporary plugging fracturing (TPF) process, the pressure response and pumping behavior significantly differ from those observed during conventional fracturing fluid pumping. Once the temporary plugging agent (TPA) forms a plug, subsequent fracture initiation and propagation become more intricate due to the influence of the TPA and early fractures. Factors such as concentration, particle size, and ratio of the TPA notably affect the effectiveness of TPF. This study employs a true triaxial hydraulic fracturing simulation system to conduct TPF experiments with varying particle size combinations and concentrations at both in-fracture and in-stage locations. The impact of different TPA parameters on the plugging effectiveness is assessed by analyzing the morphology of the induced fractures and the characteristics of pressure curves post experiment. Results indicate that combining dfferent particle sizes enhances plugging effectiveness, with a combination of smaller and larger particles exhibiting superior plugging effectiveness, resulting in a pressure increase of over 25.9%. As the concentration of the TPA increases, the plugging fracture pressure rises, accompanied by rapid pressure response and significant plugging effects, leading to more complex fracture morphology. For shale reservoirs, the density of bedding planes (BPs) influences the morphology and width of conventional hydraulic fractures, thereby affecting the effectiveness of subsequent refracturing. Rock samples with a relatively low BP density demonstrate effective plugging initiation both in-fracture and in-stage, facilitating the formation of complex fracture networks. Conversely, specimens with a relatively high BP density exhibit superior plugging effectiveness in-stage compared to in-fracture plugging. Full article

Background: Age-related macular degeneration (AMD) is a progressive ocular ailment causing age-associated vision deterioration, characterized by dysregulated immune cell activity. Notably, follicular helper T (Tfh) cells have emerged as pivotal contributors to AMD pathogenesis. Nonetheless, investigations into Tfh-associated gene biomarkers for this disorder remain limited. Methods: Utilizing gene expression data pertinent to AMD procured from the Gene Expression Omnibus (GEO) repository, we employed the “DESeq2” R software package to standardize and preprocess expression levels. Concurrently, CIBERSORT analysis was utilized to compute the infiltration proportions of 22 distinct immune cell types. Subsequent to weighted gene correlation network analysis (WGCNA), coupled with differential expression scrutiny, we pinpointed genes intricately linked with Tfh cells. These potential genes underwent further screening using the MCODE function within Cytoscape software. Ultimately, a judicious selection of pivotal genes from these identified clusters was executed through the LASSO algorithm. Subsequently, a diagnostic nomogram was devised based on these selected genes. Results: Evident Tfh cell disparities between AMD and control cohorts were observed. Our amalgamated analysis, amalgamating differential expression data with co-expression patterns, unveiled six genes closely associated with Tfh cells in AMD. Subsequent employment of the LASSO algo-rithm facilitated identification of the most pertinent genes conducive to predictive modeling. From these, GABRB3, MFF, and PROX1 were elected as prospective diagnostic biomarkers for AMD. Conclusions: This investigation discerned three novel biomarker genes, linked to inflammatory mechanisms and pivotal in diagnosing AMD. Further exploration of these genes holds potential to foster novel therapeutic modalities and augment comprehension of AMD’s disease trajectory. Full article