Search Results (339)

Weed detection is critical for precision agriculture, enabling targeted herbicide application to reduce costs and enhance crop health. This study utilized UAV-acquired RGB imagery from cotton fields to develop and evaluate deep learning models for weed detection. As sustainable resource management gains importance in rainfed agricultural systems, precise weed identification is essential to optimize yields and minimize herbicide use. However, distinguishing weeds from crops in complex field environments remains challenging due to their visual similarity. This research employed YOLOv7, YOLOv7-w6, and YOLOv7-x models to detect and classify weeds in cotton fields, using a dataset of 9249 images collected under real field conditions. To improve model performance, we enhanced the annotation process using LabelImg and Roboflow, ensuring accurate separation of weeds and cotton plants. Additionally, we fine-tuned key hyperparameters, including batch size, epochs, and input resolution, to optimize detection performance. YOLOv7, achieving the highest estimated accuracy at 83%, demonstrated superior weed detection sensitivity, particularly in cluttered field conditions, while YOLOv7-x with accuracy at 77% offered balanced performance across both cotton and weed classes. YOLOv7-w6 with accuracy at 63% faced difficulties in distinguishing features in shaded or cluttered soil regions. These findings highlight the potential of UAV-based deep learning approaches to support site-specific weed management in cotton fields, providing an efficient, environmentally friendly approach to weed management. Full article

Eel populations are globally threatened by overfishing and illegal trade, making accurate species identification essential for resource conservation and regulatory enforcement. Conventional molecular identification methods are generally applied in the laboratory, with limited rapid on-site application. This study developed a field-deployable assay to identify the Japanese eel (Anguilla japonica), by incorporating multienzyme isothermal rapid amplification (MIRA) technology with a visually readable lateral flow assay (LFA). Species-specific primers targeting a 286 bp region within the mitochondrial genome of A. japonica were designed and labeled with fluorescein amidite and biotin, respectively. The performance of the MIRA-LFA was validated by assessing its specificity against four other major eel species and its analytical sensitivity, i.e., limit of detection (LoD), under optimized temperature and reaction-time conditions. The MIRA-LFA demonstrated 100% specificity, generating a positive signal only for A. japonica, with no cross-reactivity. A clear visual result was obtained within 10 min at the optimal reaction temperature of 39 °C. Under these optimal conditions, the assay showed a high sensitivity, with an LoD of 0.1 ng/μL of genomic DNA. The proposed assay is an effective tool for the rapid, specific, and sensitive identification of A. japonica. The ability to obtain fast, equipment-free visual results makes this assay an ideal point-of-care testing solution to combat seafood fraud and support the sustainable management of this economically important and vulnerable species. Full article

Apurinic/apyrimidinic endonuclease 1 (APE1) selectively cleaves the apurinic/apyrimidinic site (AP site) in DNA, playing a critical role in base excision repair and genomic stability maintenance. Aberrant APE1 expression has been linked to numerous diseases, including cardiovascular disorders, neurological conditions, and various cancers. However, existing methods for detecting trace levels of APE1 remain suboptimal for certain applications. To address this limitation, we developed an innovative biosensing platform for ultrasensitive APE1 detection by integrating APE1-specific site recognition with hybridization chain reaction (HCR)-based signal amplification, enabling enzyme- and label-free bioassays. In this system, APE1 recognizes and cleaves the AP site-containing hairpin (HP) probe, releasing a single-stranded HCR initiator that triggers cascaded HCR amplification. Owing to the high efficiency of HCR, this method achieves ultrahigh sensitivity, with a calculated detection limit of 1.0 × 10⁻⁸ U/mL. Furthermore, the biosensor demonstrates robust performance in cell lysates and is applicable for screening and evaluating APE1 inhibitors. Full article

The N-glycolylneuraminic acid (Neu5Gc), a major salivary acid molecule found on the cell surface of animals such as pigs, cows, and sheep, can be metabolically incorporated into the body through consumption of animal-derived foods like red meat. This leads to an immune response and chronic inflammation in individuals who do not naturally produce Neu5Gc, including humans and poultry, further increasing the risk of cancer. The trans-cleavage activity of Cas12a is activated by the recognition of the target aptamer by the crRNA, resulting in the cleavage of the dual-labeled probe. By combining this with immunochromatographic techniques, we established a chromatographic test strip assay that allows immediate on-site detection of Neu5Gc contamination in non-red meat samples devoid of Neu5Gc. Further optimization enabled specific detection within 25 min with a minimum detectable limit of 10 ng/mL. These analyses successfully detected the spiked samples and actual samples containing Neu5Gc. The developed lateral flow test strips based on aptamer-Cas12a can be utilized for detecting Neu5Gc contamination in non-red meat food products, animal bioproducts, and poultry feeds. Full article

Early and accurate detection of plant diseases is critical for ensuring global food security and agricultural resilience. Ratoon stunting disease (RSD), caused by the bacterium Leifsonia xyli subsp. xyli (Lxx), is among the most economically significant diseases of sugarcane worldwide. Its cryptic nature—characterized by an absence of visible symptoms—renders timely diagnosis particularly difficult, contributing to substantial undetected yield losses across major sugar-producing regions. Here, we report the development of a potential-induced electrochemical (EC) nanobiosensor platform for the rapid, low-cost, and field-deployable detection of Lxx DNA directly from crude sugarcane sap. This method eliminates the need for conventional nucleic acid extraction and thermal cycling by integrating the following: (i) a boiling lysis-based DNA release from xylem sap; (ii) sequence-specific magnetic bead-based purification of Lxx DNA using immobilized capture probes; and (iii) label-free electrochemical detection using a potential-driven DNA adsorption sensing platform. The biosensor shows exceptional analytical performance, achieving a detection limit of 10 cells/µL with a broad dynamic range spanning from 10⁵ to 1 copy/µL (r = 0.99) and high reproducibility (SD < 5%, n = 3). Field validation using genetically diverse sugarcane cultivars from an inoculated trial demonstrated a strong correlation between biosensor signals and known disease resistance ratings. Quantitative results from the EC biosensor also showed a robust correlation with qPCR data (r = 0.84, n = 10, p < 0.001), confirming diagnostic accuracy. This first-in-class EC nanobiosensor for RSD represents a major technological advance over existing methods by offering a cost-effective, equipment-free, and scalable solution suitable for on-site deployment by non-specialist users. Beyond sugarcane, the modular nature of this detection platform opens up opportunities for multiplexed detection of plant pathogens, making it a transformative tool for early disease surveillance, precision agriculture, and biosecurity monitoring. This work lays the foundation for the development of a universal point-of-care platform for managing plant and crop diseases, supporting sustainable agriculture and global food resilience in the face of climate and pathogen threats. Full article

The increasing frequency of structural failures on construction sites emphasizes the critical role of rigorous supervision in ensuring the quality of both construction processes and materials. Current regulatory frameworks mandate the production of detailed supervision reports to provide comprehensive evaluations of construction quality, material compliance, and site records. This study proposes a novel approach to harnessing unstructured reports for automated quality assessment. Employing text mining techniques, a sentiment lexicon specifically tailored for quality performance evaluation was developed. A corpus-based manual classification was conducted on 291 relevant words and 432 sentences extracted from the supervision reports, assigning sentiment labels of negative, neutral, and positive. This sentiment lexicon was then utilized as fundamental information for the Quality Performance Evaluation Model (QPEM). To validate the efficacy of the QPEM, it was applied to supervision reports from 30 construction sites adhering to legal standards. Furthermore, a Pearson correlation analysis was performed with the actual outcomes based on the legal requirements, including quality test failure rate, material inspection failure rate, and inspection management performance. By leveraging the wealth of unstructured data continuously generated throughout a project’s lifecycle, this model can enhance the timeliness of inspection and management processes, ultimately contributing to improved construction performance. Full article

Glycated hemoglobin (HbA1c) is an important biomarker applied for the diagnosis, evaluation, and management of diabetes; therefore, its accurate determination is crucial. In this study, an innovative nanoplatform was developed, integrating carbon nanotubes (CNTs) with enhanced hydrophilicity achieved through cyclodextrin (CD) functionalization, and combined with gold nanoparticles (AuNPs) electrochemically deposited onto a screen-printed carbon electrode. The nanomaterials significantly improved the analytical performance of the sensor due to their increased surface area and high electrical conductivity. This nanoplatform was employed as a substrate for the covalent attachment of thiolated ferrocene-labeled HbA1c specific aptamer through Au-S binding. The electrochemical signal of ferrocene was covered by a stronger oxidation peak of Fe²⁺ from the HbA1c structure, leading to the elaboration of a nanostructured aptasensor capable of the direct detection of HbA1c. The electrochemical aptasensor presented a very wide linear range (0.688–11.5%), an acceptable limit of detection (0.098%), and good selectivity and stability, being successfully applied on real samples. This miniaturized, simple, easy-to-use, and fast-responding aptasensor, requiring only a small sample volume, can be considered as a promising candidate for the efficient on-site determination of HbA1c. Full article

The integration of unmanned aerial vehicle (UAV) remote sensing and deep learning has emerged as a highly effective strategy for inventorying forest resources. However, the spatiotemporal variability of forest environments and the scarcity of annotated data hinder the performance of conventional supervised deep-learning models. To overcome these challenges, this study has developed efficient tree (ET), a semi-supervised tree detector designed for forest scenes. ET employed an enhanced YOLO model (YOLO-Tree) as a base detector and incorporated a teacher–student semi-supervised learning (SSL) framework based on pseudo-labeling, effectively leveraging abundant unlabeled data to bolster model robustness. The results revealed that SSL significantly improved outcomes in scenarios with sparse labeled data, specifically when the annotation proportion was below 50%. Additionally, employing overlapping cropping as a data augmentation strategy mitigated instability during semi-supervised training under conditions of limited sample size. Notably, introducing unlabeled data from external sites enhances the accuracy and cross-site generalization of models trained on diverse datasets, achieving impressive results with F1, mAP50, and mAP50-95 scores of 0.979, 0.992, and 0.871, respectively. In conclusion, this study highlights the potential of combining UAV-based RGB imagery with SSL to advance tree species identification in heterogeneous forests. Full article

We propose a data-efficient framework for automating sowing operations by agricultural robots in densely mixed polyculture environments. This study addresses the challenge of enabling robots to identify suitable sowing positions with minimal labeled data by integrating image-based field sensing with expert agricultural knowledge. We collected 84 RGB-depth images from seven field sites, labeled by synecological farming practitioners of varying proficiency levels, and trained a regression model to estimate optimal sowing positions and seeding quantities. The model’s predictions were comparable to those of intermediate-to-advanced practitioners across diverse field conditions. To implement this estimation in practice, we mounted a Kinect v2 sensor on a robot arm and integrated its 3D spatial data with axis-specific movement control. We then applied a trajectory optimization algorithm based on the traveling salesman problem to generate efficient sowing paths. Simulated trials incorporating both computation and robotic control times showed that our method reduced sowing operation time by 51% compared to random planning. These findings highlight the potential of interpretable, low-data machine learning models for rapid adaptation to complex agroecological systems and demonstrate a practical approach to combining structured human expertise with sensor-based automation in biodiverse farming environments. Full article

Smart surveillance cameras are increasingly employed for automated tasks such as event and anomaly detection within smart city infrastructures. However, the heterogeneity of deployment environments, ranging from densely populated urban intersections to quiet residential neighborhoods, renders the use of a single, universal model suboptimal. To address this, we propose the construction of a model zoo comprising models trained for diverse environmental contexts. We introduce an automated transferability assessment framework that identifies the most suitable model for a new deployment site. This task is particularly challenging in smart surveillance settings, where both source data and labeled target data are typically unavailable. Existing approaches often depend on pretrained embeddings or assumptions about model uncertainty, which may not hold reliably in real-world scenarios. In contrast, our method leverages embeddings generated by randomly initialized neural networks (RINNs) to construct task-agnostic reference embeddings without relying on pretraining. By comparing feature representations of the target data extracted using both pretrained models and RINNs, this method eliminates the need for labeled data. Structural similarity between embeddings is quantified using minibatch-Centered Kernel Alignment (CKA), enabling efficient and scalable model ranking. We evaluate our method on realistic surveillance datasets across multiple downstream tasks, including object tagging, anomaly detection, and event classification. Our embedding-level score achieves high correlations with ground-truth model rankings (relative to fine-tuned baselines), attaining Kendall’s $\tau$ values of 0.95, 0.94, and 0.89 on these tasks, respectively. These results demonstrate that our framework consistently selects the most transferable model, even when the specific downstream task or objective is unknown. This confirms the practicality of our approach as a robust, low-cost precursor to model adaptation or retraining. Full article

Autoantibodies targeting cytosolic 5′-nucleotidase 1A (cN1A) are found in several autoimmune diseases, including inclusion body myositis (IBM), Sjögren’s syndrome, and systemic lupus erythematosus. While they have diagnostic relevance for IBM, little is known about the autoreactive B cells that produce these antibodies. To address this, we developed a robust protocol for the expression and site-specific biotinylation of recombinant human cN1A in Escherichia coli. The resulting antigen is suitable for generating double-labelled fluorescent baits for the isolation and characterisation of cN1A-specific B cells by flow cytometry. Site-specific biotinylation was achieved using the AviTag and BirA ligase, preserving the protein’s structure and immunoreactivity. Western blot analysis confirmed that the biotinylated cN1A was recognised by both human and rabbit anti-cN1A antibodies. Compared to conventional chemical biotinylation, this strategy minimises structural alterations that may affect antigen recognition. This approach provides a reliable method for producing biotinylated antigens for use in immunological assays. While demonstrated here for cN1A, the protocol can be adapted for other autoantigens to support studies of antigen-specific B cells in autoimmune diseases. Full article

Zearalenone (ZEA), a potent mycotoxin commonly found in contaminated grains, presents a serious threat to food safety and public health. Conventional detection methods, including culture-based assays and laboratory-bound analytical tools, are often time-consuming, require specialized infrastructure, and lack portability, limiting their utility for rapid, on-site screening. In response, this study introduces a compact, real-time electrochemical sensing platform for the swift and selective detection of ZEA in corn flour matrices. Utilizing a non-faradaic, label-free approach based on Electrochemical Impedance Spectroscopy (EIS), the sensor leverages ZEA-specific antibodies to achieve rapid detection within 5 min. The platform demonstrates a low detection limit of 0.05 ng/mL, with a broad dynamic range from 0.1 ng/mL to 25.6 ng/mL. Reproducibility tests confirm consistent performance, with both inter- and intra-assay variation remaining under a 20% coefficient of variation (%CV). Comparative evaluation with standard benchtop systems underscores its accuracy and field applicability. This portable and user-friendly device provides a powerful tool for real-time mycotoxin monitoring, offering significant potential for improving food safety practices and enabling point-of-need testing in resource-limited settings. Full article

The immunofluorescence assay is widely used for cellular biology and diagnosis applications. Such an antigen–antibody detection system enables the assessment and visualization of the expression and localization of target proteins. In the classical indirect immunofluorescence assay, secondary antibodies are conjugated to fluorophores. However, conventional secondary antibodies have limited applications due to their large size (150 kDa). Moreover, as animal-derived products, secondary antibodies are associated with ethical concerns and batch-to-batch variability. In this study, we developed fluorescence-labeled recombinant nanobodies as secondary antibodies by utilizing previously established anti–mouse and anti–rabbit IgG secondary nanobodies in combination with the self-labeling SNAP-tag. Nanobodies, which are significantly smaller (15 kDa), are capable to detect primary antibodies produced in mice and rabbits. The SNAP-tag (20 kDa) enables site-specific binding of various O⁶-benzylguanine (BG)-modified fluorophores to the recombinant nanobodies. These recombinant nanobodies were produced using mammalian cell expression system, and their specific binding to mouse or rabbit antibodies was validated using flow cytometry and multi-color fluorescence microscopy. The low cost, easy of expression, purification and site-specific conjugation procedures for these anti–mouse and anti–rabbit IgG secondary nanobodies make them an attractive alternative to traditional secondary antibodies for indirect immunofluorescence assays. Full article

Background: While sentinel lymph node (SLN) biopsy has been integrated into international guidelines for endometrial cancer, a standardized technique is still lacking. This study addresses whether the concurrent use of two tracers, technetium-99 (Tc) and indocyanine green (ICG), administered intracervically through distinct techniques, enhances the performance of SLN biopsies. As the blue dye is used routinely by some centers, it can be used alone; however, our analysis focused on only Tc and ICG (as is used in the majority of centers). Methods: A prospective multicentric observational study was designed to evaluate the unilateral detection rate, bilateral detection rates, sensitivity, and consistency of SLNs when using both tracers simultaneously in patients with early-stage endometrial cancer. Results: Our findings demonstrated that the simultaneous use of ICG and Tc significantly outperformed the use of either tracer alone. Unilateral detection rates were 69.2% for Tc, 84.9% for ICG, and 88.4% for both. Bilateral detection rates were 57.0% for Tc, 77.9% for ICG, and 83.6% for both. Additionally, the incidence of “empty pockets” was low with both tracers, at 2.7%. Notably, the concurrent application of both tracers identified instances where the Tc-labeled sentinel node differed from the ICG-labeled sentinel node. Conclusions: The combined use of Tc and ICG in SLN biopsy for early-stage endometrial cancer significantly enhances detection rates and reduces the occurrence of “empty pockets”, potentially decreasing the need for site-specific lymphadenectomy. Full article

Precision livestock farming technologies support health monitoring on farms, yet few studies have evaluated their effectiveness under field conditions using reliable gold standards. This study evaluated a commercially available technology for detecting cough sounds in pigs on a commercial farm. Audio was recorded over six days using 16 microphones across two pig barns. A total of 1110 cough sounds were labelled by an on-site observer using a cough induction methodology, and 8938 other sounds from farm recordings and open-source datasets (ESC-50, UrbanSound8K, and AudioSet) were labelled. A hybrid deep learning model combining Convolutional Neural Networks and Recurrent Neural Networks was trained and evaluated using these labels. A total of 34 audio features were extracted from 1 s segments, including validated descriptors (e.g., MFCC), unverified external features, and proprietary features. Features were evaluated through 10-fold cross-validation based on classification performance and runtime, resulting in eight final features. The final model showed high performance (recall = 98.6%, specificity = 99.7%, precision = 98.8%, accuracy = 99.6%, F1-score = 98.6%). The technology tested was shown to be efficient for monitoring cough sounds in a commercial swine production facility. It is recommended to test the technology in other environments to evaluate the effectiveness in different farm settings. Full article