Search Results (3,057)

As large language models (LLMs) gain traction among researchers and practitioners, particularly in digital marketing for tasks such as customer feedback analysis and automated communication, concerns remain about the reliability and consistency of their outputs. This study investigates annotation bias in LLMs by comparing human and AI-generated annotation labels across sentiment, topic, and aspect dimensions in hotel booking reviews. Using the HRAST dataset, which includes 23,114 real user-generated review sentences and a synthetically generated corpus of 2000 LLM-authored sentences, we evaluate inter-annotator agreement between a human expert and three LLMs (ChatGPT-3.5, ChatGPT-4, and ChatGPT-4-mini) as a proxy for assessing annotation bias. Our findings show high agreement among LLMs, especially on synthetic data, but only moderate to fair alignment with human annotations, particularly in sentiment and aspect-based sentiment analysis. LLMs display a pronounced neutrality bias, often defaulting to neutral sentiment in ambiguous cases. Moreover, annotation behavior varies notably with task design, as manual, one-to-one prompting produces higher agreement with human labels than automated batch processing. The study identifies three distinct AI biases—repetition bias, behavioral bias, and neutrality bias—that shape annotation outcomes. These findings highlight how dataset complexity and annotation mode influence LLM behavior, offering important theoretical, methodological, and practical implications for AI-assisted annotation and synthetic content generation. Full article

Direct seeding has gained prominence as a labor-efficient and environmentally sustainable alternative to conventional transplanting in rice cultivation. In direct seeding systems, early-stage management is crucial for stable seedling establishment, with sowing uniformity measured by seed counts being a critical indicator of success. However, conventional manual seed counting methods are time-consuming, prone to human error, and impractical for large-scale or repetitive tasks, necessitating advanced automated solutions. Recent advances in computer vision technologies and precision agriculture tools, offer the potential to automate seed counting tasks. Nevertheless, challenges such as domain discrepancies and limited labeled data restrict robust real-world deployment. To address these issues, we propose a density estimation-based seed counting framework integrating semi-supervised learning and background augmentation. This framework includes a cost-effective data acquisition system enabling diverse domain data collection through indoor background augmentation, combined with semi-supervised learning to utilize augmented data effectively while minimizing labeling costs. The experimental results on field data from unknown domains show that our approach reduces seed counting errors by up to 58.5% compared to conventional methods, highlighting its potential as a scalable and effective solution for agricultural applications in real-world environments. Full article

This research presents the development of a novel object detection algorithm designed to identify humans in natural outdoor environments using minimal computational resources. The proposed system, SARGAS, combines a custom convolutional neural network (CNN) classifier with MOG2 background subtraction and partial affine transformations for camera stabilization. A secondary CNN refines detections and reduces false positives. Unlike conventional supervised models, SARGAS is trained in a partially unsupervised manner, learning to recognize feature patterns without requiring labeled data. The algorithm achieved a recall of 93%, demonstrating strong detection capability even under challenging conditions. However, the overall accuracy reached 65%, due to a higher rate of false positives—an expected trade-off when maximizing recall. This bias is intentional, as missing a human target in search and rescue applications carries a higher cost than producing additional false detections. While supervised models, such as YOLOv5, perform well on data resembling their training sets, they exhibit significant performance degradation on previously unseen footage. In contrast, SARGAS generalizes more effectively, making it a promising candidate for real-world deployment in environments where labeled training data is limited or unavailable. The results establish a solid foundation for further improvements and suggest that unsupervised CNN-based approaches hold strong potential in object detection tasks. Full article

Super-resolution microscopy (SRM) has revolutionized our understanding of subcellular structures, including cell organelles and viruses. For human immunodeficiency virus (HIV), SRM has significantly advanced knowledge of viral structural biology and assembly dynamics. This review analyzes how SRM techniques (particularly PALM, STORM, STED, and SIM) have been applied over the past decade to study HIV structural components and assembly. By categorizing and comparing studies based on SRM methods, HIV components, and labeling strategies, we assess the strengths and limitations of each approach. Our analysis shows that PALM is most commonly used for live-cell imaging of HIV Gag, while STED is primarily used to study the viral envelope (Env). STORM and SIM have been applied to visualize various components, including Env, capsid, and matrix. Antibody labeling is prevalent in PALM and STORM studies, targeting Env and capsid, whereas fluorescent protein labeling is mainly associated with PALM and focused on Gag. A recent emphasis on Gag and Env points to deeper investigation into HIV assembly and viral membrane dynamics. Insights from SRM studies of HIV not only enhance virological understanding but also inform future research in therapeutic strategies and delivery systems, including extracellular vesicles. Full article

Accounting for 15–30% of colorectal cancer cases, the serrated pathway remains poorly characterized compared to the adenoma–carcinoma sequence. It involves sessile serrated lesions as precursors and is characterized by BRAF mutations (BRAF^V600E), CpG island hypermethylation, and microsatellite instability (MSI). Using label-free proteomics, we compared normal tissue margins from patients with diverticular disease, sessile serrated lesions, low-grade adenomas, and high-grade adenomas. We identified S100A14 as significantly overexpressed in sessile serrated lesions compared to low-grade adenomas, high-grade adenomas, and normal tissues. This overexpression was confirmed by immunohistochemical scoring in an independent cohort. Gene expression analyses of public datasets showed higher S100A14 expression in BRAF^V600E-mutated and MSI-H colorectal cancers compared to microsatellite stable BRAF^wt tumors. This finding was confirmed by immunohistochemical scoring in an independent colorectal cancer cohort. Furthermore, single-cell RNA sequencing analysis from the Human Colon Cancer Atlas revealed that S100A14 expression in tumor cells positively correlated with the abundance of tumoral CD8⁺ cytotoxic T cells, particularly the CD8⁺ CXCL13⁺ subset, known for its association with a favorable response to immunotherapy. Collectively, our results demonstrate for the first time that S100A14 is a potential biomarker of serrated neoplasia and further suggests its potential role in predicting immunotherapy responses in colorectal cancer. Full article

Due to their nutritional value and sustainability, edible insect-based foods are gaining popularity in Europe. Their use is regulated by EU legislation, which defines authorised species and sets labelling requirements. Molecular tools are being developed to authenticate such products. In this study, yellow mealworm (Tenebrio molitor) larvae authorised for human consumption were fed wheat flour-based diets containing varying proportions of house cricket (Acheta domesticus) flour for 21 days. This was followed by a 48 h starvation period to assess the persistence of insect DNA in the digestive tract. Two novel, species-specific, single-copy markers were designed: ampd gene for the Acheta domesticus and MyD88 gene for the Tenebrio molitor. These were applied using qPCR and ddPCR. Both methods successfully detected cricket DNA in the guts of starved larvae. Linear regression analysis revealed a strong, statistically significant correlation between the proportion of Acheta domesticus flour in the diet and the normalised relative quantity of DNA. ddPCR proved to be more sensitive than qPCR, particularly in the detection of low DNA levels. These results suggest that the presence of DNA from undeclared insect species in edible insects may be indicative of their diet rather than contamination or adulteration. This highlights the importance of contextual interpretation in food authenticity testing. Full article

Pompe disease is a rare autosomal-recessive neuromuscular disorder caused by a deficiency of the lysosomal enzyme acid alpha-glucosidase (GAA), leading to the pathological accumulation of glycogen and impaired autophagy. Enzyme replacement therapy (ERT) with recombinant human alpha-glucosidase (rhGAA) has been available since 2006, but may lead to the formation of anti-drug antibodies (ADAs) against the recombinant human enzyme, which, in turn, may adversely affect the response to ERT. Knowledge of the antigenic determinants of rhGAA involved in interaction with ADAs may facilitate the development of strategies to attenuate the anti-drug immune response in patients. Here, we determined the rhGAA ADA epitopes in the plasma of Pompe disease patients using a series of affinity purifications combined with epitope extraction and label free quantitation LC-MS. Full article

Robotic search of people in human-centered environments, including healthcare settings, is challenging, as autonomous robots need to locate people without complete or any prior knowledge of their schedules, plans, or locations. Furthermore, robots need to be able to adapt to real-time events that can influence a person’s plan in an environment. In this paper, we present MLLM-Search, a novel zero-shot person search architecture that leverages multimodal large language models (MLLM) to address the mobile robot problem of searching for a person under event-driven scenarios with varying user schedules. Our approach introduces a novel visual prompting method to provide robots with spatial understanding of the environment by generating a spatially grounded waypoint map, representing navigable waypoints using a topological graph and regions by semantic labels. This is incorporated into an MLLM with a region planner that selects the next search region based on the semantic relevance to the search scenario and a waypoint planner that generates a search path by considering the semantically relevant objects and the local spatial context through our unique spatial chain-of-thought prompting approach. Extensive 3D photorealistic experiments were conducted to validate the performance of MLLM-Search in searching for a person with a changing schedule in different environments. An ablation study was also conducted to validate the main design choices of MLLM-Search. Furthermore, a comparison study with state-of-the-art search methods demonstrated that MLLM-Search outperforms existing methods with respect to search efficiency. Real-world experiments with a mobile robot in a multi-room floor of a building showed that MLLM-Search was able to generalize to new and unseen environments. Full article

Occlusion remains a major challenge in person re-identification, as it often leads to incomplete or misleading visual cues. To address this issue, we propose a dual-branch occlusion-aware network (DOAN), which explicitly and implicitly enhances the model’s capability to perceive and handle occlusions. The proposed DOAN framework comprises two synergistic branches. In the first branch, we introduce an Occlusion-Aware Semantic Attention (OASA) module to extract semantic part features, incorporating a parallel channel and spatial attention (PCSA) block to precisely distinguish between pedestrian body regions and occlusion noise. We also generate occlusion-aware parsing labels by combining external human parsing annotations with occluder masks, providing structural supervision to guide the model in focusing on visible regions. In the second branch, we develop an occlusion-aware recovery (OAR) module that reconstructs occluded pedestrians to their original, unoccluded form, enabling the model to recover missing semantic information and enhance occlusion robustness. Extensive experiments on occluded, partial, and holistic benchmark datasets demonstrate that DOAN consistently outperforms existing state-of-the-art methods. Full article

Background: Lysosomal-associated membrane protein 1 (LAMP1), typically localized to the lysosomal membrane, is increasingly implicated as a marker of cancer aggressiveness and metastasis when expressed on the cell surface. This study aimed to develop a LAMP1-targeted antibody-based PET tracer and assess its efficacy in mouse models of human breast and colon adenocarcinoma. Methods: To determine the source of LAMP1 expression, we utilized human single-cell RNA sequencing and spatial transcriptomics, complemented by in-house flow cytometry on xenografted mouse models. Tissue microarrays of multiple epithelial cancers and normal tissue were stained for LAMP-1, and staining was quantified. An anti-LAMP1 monoclonal antibody was conjugated with desferrioxamine (DFO) and labeled with zirconium-89 (⁸⁹Zr). Human triple-negative breast cancer (MDA-MB-231) and colon cancer (Caco-2) cell lines were implanted in nude mice. PET/CT imaging was conducted at 24, 72, and 168 h post-intravenous injection of ⁸⁹Zr-DFO-anti-LAMP1 and ⁸⁹Zr-DFO-IgG (negative control), followed by organ-specific biodistribution analyses at the final imaging time point. Results: Integrated single-cell and spatial RNA sequencing demonstrated that LAMP1 expression was localized to myeloid-derived suppressor cells (MDSCs) and cancer-associated fibroblasts (CAFs) in addition to the cancer cells. Tissue microarray showed significantly higher staining for LAMP-1 in tumor tissue compared to normal tissue (3986 ± 2635 vs. 1299 ± 1291, p < 0.001). Additionally, xenograft models showed a significantly higher contribution of cancer cells than the immune cells to cell surface LAMP1 expression. In vivo, PET imaging with ⁸⁹Zr-DFO-anti-LAMP1 PET/CT revealed detectable tumor uptake as early as 24 h post-injection. The ⁸⁹Zr-DFO-anti-LAMP1 tracer demonstrated significantly higher uptake than the control ⁸⁹Zr-DFO-IgG in both models across all time points (MDA-MB-231 SUV_max at 168 h: 12.9 ± 5.7 vs. 4.4 ± 2.4, p = 0.003; Caco-2 SUV_max at 168 h: 8.53 ± 3.03 vs. 3.38 ± 1.25, p < 0.01). Conclusions: Imaging of cell surface LAMP-1 in breast and colon adenocarcinoma is feasible by immuno-PET. LAMP-1 imaging can be expanded to adenocarcinomas of other origins, such as prostate and pancreas. Full article

Using neural networks has become the standard technique for medical diagnostics, especially in cancer detection and classification. This work evaluates the performance of Vision Transformer architectures, including Swin Transformer and MaxViT, for several datasets of magnetic resonance imaging (MRI) and computed tomography (CT) scans. We used three training sets of images with brain, lung, and kidney tumors. Each dataset included different classification labels, from brain gliomas and meningiomas to benign and malignant lung conditions and kidney anomalies such as cysts and cancers. This work aims to analyze the behavior of the neural networks in each dataset and the benefits of combining different image modalities and tumor classes. We designed several experiments by fine-tuning the models on combined and individual datasets. The results revealed that the Swin Transformer achieved the highest accuracy, with an average of 99.0% on single datasets and reaching 99.43% on the combined dataset. This research highlights the adaptability of Transformer-based models to various human organs and image modalities. The main contribution lies in evaluating multiple ViT architectures across multi-organ tumor datasets, demonstrating their generalization to multi-organ classification. Integrating these models across diverse datasets could mark a significant advance in precision medicine, paving the way for more efficient healthcare solutions. Full article

The growing demand for accurate and efficient Chest X-Ray (CXR) interpretation has prompted the development of AI-driven systems to alleviate radiologist workload and reduce diagnostic variability. This paper introduces the Intelligent Humanized Radiology Analysis System (IHRAS), a modular framework that automates the end-to-end process of CXR analysis and report generation. IHRAS integrates four core components: (i) deep convolutional neural networks for multi-label classification of 14 thoracic conditions; (ii) Grad-CAM for spatial visualization of pathologies; (iii) SAR-Net for anatomical segmentation; and (iv) a large language model (DeepSeek-R1) guided by the CRISPE prompt engineering framework to generate structured diagnostic reports using SNOMED CT terminology. Evaluated on the NIH ChestX-ray dataset, IHRAS demonstrates consistent diagnostic performance across diverse demographic and clinical subgroups, and produces high-fidelity, clinically relevant radiological reports with strong faithfulness, relevancy, and alignment scores. The system offers a transparent and scalable solution to support radiological workflows while highlighting the importance of interpretability and standardization in clinical Artificial Intelligence applications. Full article

Current mainstream classification methods predominantly employ end-to-end multi-class frameworks. These approaches encounter inherent challenges including high-dimensional feature space complexity, decision boundary ambiguity that escalates with increasing class cardinality, sensitivity to label noise, and limited adaptability to dynamic model expansion. However, human beings may avoid these mistakes naturally. Research indicates that humans subconsciously employ a decision-making process favoring binary outcomes, particularly when responding to questions requiring nuanced differentiation. Intuitively, responding to binary inquiries such as “yes/no” often proves easier for humans than addressing queries of “what/which”. Inspired by the human decision-making hypothesis, we proposes a decision paradigm named the evolutionary binary decision framework (EBDF) centered around binary classification, evolving from traditional multi-classifiers in deep learning. To facilitate this evolution, we leverage the top-N outputs from the traditional multi-class classifier to dynamically steer subsequent binary classifiers, thereby constructing a cascaded decision-making framework that emulates the hierarchical reasoning of a binary decision tree. Theoretically, we demonstrate mathematical proof that by surpassing a certain threshold of the performance of binary classifiers, our framework may outperform traditional multi-classification framework. Furthermore, we conduct experiments utilizing several prominent deep learning models across various image classification datasets. The experimental results indicate significant potential for our strategy to surpass the ceiling in multi-classification performance. Full article

Modified oligonucleotides (oligos) are widely used as convenient tools in many scientific fields, including biomedical applications and therapies. In particular, oligos with lipophilic groups attached to the backbone ensure penetration of the cell membrane without the need for transfection. This study examines the interaction between amphiphilic DNA duplexes, in which one of the chains contains a lipophilic substituent, and several DNA repair proteins, particularly DNA-damage-dependent PARPs, using various biochemical approaches. DNA with a lipophilic substituent (LS-DNA) demonstrates more efficient binding with DNA damage activated poly(AD-ribose) polymerases 1-3 (PARP1, PARP2, PARP3) and DNA polymerase β. Chemically reactive LS-DNA derivatives containing a photoactivatable nucleotide (photo-LS-DNAs) or a 5′ deoxyribose phosphate (dRP) group in the vicinity of double-strand breaks (DSBs) are used for the affinity labelling of PARPs and other proteins in several whole-cell extracts of human cells. In particular, photo-LS-DNAs are used to track the level of Ku antigen in the extracts of neuron-like differentiated SH-SY5Y, undifferentiated SH-SY5Y, and olfactory epithelial cells. In vitro, PARP1–PARP3 are shown to be able to slowly excise the 5′ dRP group at DSBs. LS-DNAs can activate PARP1 and PARP2 for autoPARylation, albeit less effectively than regular DNA duplexes. Full article

Estradiol (E2) plays an important role in cell proliferation and certain brain functions. To reveal its mechanism of action, its detectability is essential. Only a few fluorescent-labeled hormonally active E2s exist in the literature, and their mechanism of action usually remains unclear. It would be of particular interest to develop novel labeled estradiol derivatives with retained biological activity and improved optical properties. Due to their superior optical characteristics, aza-BODIPY dyes are frequently used labeling agents in biomedical applications. E2 was labeled with the aza-BODIPY dye at its phenolic hydroxy function via an alkyl linker and a triazole coupling moiety. The estrogenic activity of the newly synthesized fluorescent conjugate was evaluated via transcriptional luciferase assay. Docking calculations were performed for the classical and alternative binding sites (CBS and ABS) of human estrogen receptor α. The terminal alkyne function was introduced into the tetraphenyl aza-BODIPY core via selective formylation, oxidation, and subsequent amidation with propargyl amine. The conjugation was achieved via Cu(I)-catalyzed azide–alkyne click reaction of the aza-BODIPY-alkyne with the 3-O-(4-azidobut-1-yl) derivative of E2. The labeled estrogen induced a dose-dependent transcriptional activity of human estrogen receptor α with a submicromolar EC₅₀ value. Docking calculations revealed that the steroid part has a perfect overlap with E2 in ABS. In CBS, however, a head-tail binding deviation was observed. A facile, fluorescent labeling methodology has been elaborated for the development of a novel red-emitting E2 conjugate with substantial estrogenic activity. Docking experiments uncovered the binding mode of the conjugate in both ABS and CBS. Full article