Search Results (1,845)

The increasing interest in valorizing agricultural by-products has positioned grape pomace as a rich source of bioactive compounds. This study developed an effect-directed extraction (EDE) approach guided by bioactivity quantification on thin layer chromatography (TLC). Twelve grape pomaces were screened based on antioxidant and tyrosinase inhibitory properties. Using hydroalcoholic solvent (ethanol:water, 1:1), the two most promising sources (Malbec from San Rafael) were subjected to response surface methodology (RSM) to optimize extraction of anti-browning and antioxidant compounds visualized as TLC spots. Temperature and time were optimized (76 °C, 45 min), and samples were analyzed using TLC coupled with DPPH and laccase inhibition bioautography. Antioxidant compounds showed retention factor values on TLC plates of 0.37 and 0.75 (DPPH/ABTS-active), while laccase inhibition occurred at Rf 0.35, coinciding with the primary tyrosinase inhibition zone. However, subsequent bioassay-guided HPLC fractionation and HRMS/MS analysis revealed that tyrosinase and laccase inhibitions are mediated by distinct compounds within this bioactive zone, highlighting a synergistic multi-target effect in the optimized extract that is retained throughout the process. The primary tyrosinase inhibitor at Rf ~0.35 was tentatively elucidated as an acylated anthocyanin, consistent with malvidin-3-O-(p-coumaroyl)glucoside. Optimized extracts were evaluated on Pink Lady apple slices at different timepoints. The browning index was reduced by 25% versus the control at 15 h, confirmed by significantly lower ΔE values (p < 0.05). The process requires only food-grade solvents and conventional equipment, facilitating scale-up for grape pomace generated worldwide. Validating the EDE strategy, this TLC-guided approach successfully tracked and preserved the primary anti-tyrosinase activity from the crude waste matrix down to the tentatively identified molecule, contributing to circular economy objectives in the wine industry. Full article

Background: The integration of multi-omics data, such as genomics and transcriptomics, into artificial intelligence models has advanced precision medicine. However, their clinical applicability remains limited due to model complexity. We integrated DNA mutation, RNA expression, and A>I(G) RNA editing data to develop a predictive model for drug response in breast cancer. Methods: We analyzed 104 patients from the Breast Cancer Genome-Guided Therapy Study (ClinicalTrials.gov: NCT02022202). Clinical variables, gene expression, tumor and germline DNA variants, and RNA editing features were integrated into machine learning models to predict therapy response. Generalized linear models (GLM), random forest (RF), and support vector machines (SVM) were trained and evaluated across multiple random 70/30 train-test splits. Feature selection was performed exclusively within the training set using LASSO regularization. Model performance was assessed using the F1-score on independent test sets. The additive effect of RNA editing was evaluated using paired comparisons across identical train/test splits. Results: We characterized the cohort using clinical, mutational, transcriptomic, and RNA editing profiles in 69 non-responders and 35 responders. Across repeated splits, adding RNA editing frequently maintained or modestly improved predictive performance, particularly in expression-based models, with paired analyses showing a statistically significant increase in F1-score. Conclusions: RNA editing represents a complementary molecular layer that can enhance multi-omic models for therapy response prediction in breast cancer, supporting further investigation of epitranscriptomic features in precision oncology. Full article

We present an investigation of the thermal damage threshold of passivated Bi₂Se₃ films upon laser illumination, with a focus on their employment in terahertz (THz) spectroscopic applications. Passivation was achieved by depositing a thin 3 nm Al capping layer which, exposed to the ambient, forms a natural oxide. In THz transient emission experiments, the samples were exposed to a train of 100 fs wide laser pulses with 800 nm wavelength at 78 MHz repetition rate and peak power density up to 295 mW/µm². For the sake of comparison, the films were also exposed to continuous wave laser light with a wavelength of 532 nm in the average optical power density range from 5 × 10⁻² mW/µm² to 50 mW/µm². In both cases, changes in film appearance, detected by optical microscopy, or even film removal in a small area close to the center of the illuminated spot could be induced. Raman spectroscopy provided evidence that the crystalline phase of Bi₂Se₃ films is present in areas that have been exposed but not damaged. Conversely, in the film region illuminated with the highest peak power density no Raman signal was detected in the range under investigation which we ascribe to material removal. At the perimeter of this ablated area, we observed a dominant Raman mode at approximately 255 cm⁻¹ that we can attribute to selenium and indicates partial Bi₂Se₃ decomposition. In contrast, we observed Raman spectra corresponding to as-deposited Bi₂Se₃ only a few micrometers away from the laser-damaged area. Hence, the observed THz radiation originates from this illuminated but undamaged region. This detailed knowledge is expected to serve as a guide for designing the emitter’s thermal management and choosing laser parameters for optimal operation. Full article

Efficient UAV logistics in complex urban airspaces requires a synergistic approach to task allocation and path planning. However, traditional methods often decouple these two phases, leading to physically infeasible or sub-optimal delivery schedules. This paper proposes a Dual-Layer Optimization Framework (D-LOF) to address the Multi-UAV delivery problem in 3D urban environments. The upper layer utilizes an improved Genetic Algorithm (GA) with a specialized constraint repair operator to optimize task sequences for a heterogeneous UAV fleet. The lower layer employs an altitude-aware A* algorithm that dynamically balances vertical energy costs and horizontal cruise efficiency across multiple altitude layers. Unlike conventional models, our framework iteratively feeds precise 3D flight costs from the lower layer back to the upper layer to guide evolutionary search. Simulation results demonstrate that the D-LOF consistently achieves global convergence within 20 generations. Compared to single-altitude planning and rule-based strategies, the proposed method can reduce total operational costs and maintains zero time-window violations in high-density obstacle scenarios. This study provides a robust decision-making tool for “last-mile” urban logistics by navigating the trade-offs between 3D spatial constraints and delivery punctuality. Full article

Concept Bottleneck Models facilitate interpretable image classification by predicting human-understandable concepts prior to class labels. However, when constructed upon CLIP, they exhibit unreliable concept scores stemming from CLIP’s global representation bias and insufficient region-level sensitivity, which severely constrain their effectiveness in sensor-driven applications like remote sensing and medical imaging where localized visual evidence is critical. To mitigate this, we propose the Local–Global Aware Concept Bottleneck Model (LGA-CBM), which improves concept prediction through a training-free refinement pipeline. Building on initial CLIP-derived concept scores, LGA-CBM incorporates three key components: a Dual Masking Guided Concept Score Refinement (DMCSR) module that exploits attention weights to strengthen region–concept alignment; a Local-to-Global Concept Reidentification (L2GCR) strategy to harmonize local and global activations; and a Similar Concepts Correction Mechanism (SCCM) integrating Grounding DINO for fine-grained disambiguation. A sparse linear layer then maps the refined concepts to class labels, enabling highly interpretable classification with minimal concept usage. Experiments across six benchmark datasets demonstrate that LGA-CBM consistently achieves state-of-the-art performance in both accuracy and interpretability, producing explanations that align closely with human cognition. Full article

As traffic demand increases and intelligent transportation systems continue to develop, traffic signal control must operate reliably in complex and heterogeneous network environments, especially under communication instability. Traditional approaches often lack sufficient resilience when facing packet loss, delay, and other communication disturbances. This study proposes a resilient collaborative control (RCC) method for transportation hubs that explicitly considers communication reliability. A multi-layer computational framework is developed to support real-time mapping and interaction between physical and virtual networks. A fuzzy-logic-based communication state perception model is introduced to guide adaptive control-mode switching. To improve network-level performance, a recovery-oriented optimization algorithm is applied for dynamic load balancing across the hub area. Co-simulation results show that, compared with traditional adaptive control, the proposed method reduces average vehicle delay by 42.3%, increases network speed by 52.3%, shortens recovery time by 63%, and improves the resilience index to 0.87. These results support the effectiveness of the proposed framework within the evaluated co-simulation setting. Full article

Hyperspectral unmixing (HU) is fundamental for conducting quantitative analyses in remote sensing, yet existing methods face a persistent tradeoff between model performance and physical interpretability. Although deep learning models achieve superior performance, even “gray-box” models that incorporate physical constraints still suffer from an intrinsically opaque decision-making process, which hinders their trustworthiness in critical applications. To address this challenge, this paper introduces a theory-guided unmixing framework aimed at enhancing mechanistic interpretability called the sparse and subspace-attentive transformer unmixing network (SSTU-Net). Unlike heuristic architectures, SSTU-Net is rigorously derived from the first principles of sparse rate reduction (SRR) theory. Its core modules—the multi-head subspace self-attention (MSSA) and the iterative shrinkage-thresholding algorithm (ISTA)—directly implement the essential mathematical steps of information compression and sparsification within the SRR theory, respectively. Extensive experiments on both synthetic and real hyperspectral datasets demonstrate that SSTU-Net achieves competitive performance compared to representative state-of-the-art methods—including advanced autoencoder-based networks (e.g., CyCU-Net and DAAN) and recent transformer-based unmixing architectures (e.g., DeepTrans and MAT-Net)—while strictly adhering to theoretically predicted evolutionary trajectories. More importantly, a series of specifically designed structural interpretability validation experiments mechanistically confirm the theoretically predicted behaviors, such as layer-wise information compression, feature sparsification, and subspace orthogonalization. These results reveal the internal working mechanisms of SSTU-Net, validating the feasibility and significant potential of our principled theory-guided framework for developing high-performance and trustworthy intelligent models in remote sensing. Full article

Bolt-nut fasteners are critical components of substation equipment, and their integrity directly affects the operational reliability of power systems. In practical inspection scenarios, however, the small physical scale of bolt-nut fasteners, together with complex background structures, often obscures their discriminative visual features, making accurate automated detection particularly challenging. Reliable detection is a prerequisite for downstream tasks such as loosening identification and defect diagnosis. To address these challenges, this paper proposes YOLOv8n-ALC, an enhanced detection network built upon the lightweight YOLOv8n framework. The backbone is redesigned by integrating the AdditiveBlock from CAS-ViT and a Convolutional Gated Linear Unit (CGLU) to strengthen fine-grained feature extraction and suppress background interference without increasing computational burden. In addition, an improved Large Separable Kernel Attention (LSKA) module is introduced to expand the effective receptive field while maintaining efficiency, enabling more robust multi-scale feature representation. To further alleviate feature degradation of small bolt-nut fasteners in deep layers, a Context-Guided Reconstruction Feature Pyramid Network (CGRFPN) is employed in the neck to optimize cross-layer feature fusion and enhance localization accuracy. Experimental results demonstrate that YOLOv8n-ALC achieves an mAP@0.5 of 92.1%, with precision and recall of 93.5% and 87.1%, respectively, outperforming the baseline by clear margins. These results confirm the effectiveness and robustness of the proposed method for intelligent substation inspection and bolt-nut fastener condition monitoring. Full article

Low-Rank Adaptation (LoRA) has become a widely adopted parameter-efficient fine-tuning (PEFT) technique for large language models (LLMs). LoRA’s benefits stem from its light weight and modular adapters. Standard LoRA applies adapters uniformly across all Transformer layers, implicitly assuming that each layer contributes equally to task adaptation. However, LLMs are found to have internal substructures that contribute in a disproportionate manner. In this work, we provide a theoretical analysis of how LoRA weight updates are influenced by a layer’s activation magnitude. We propose Act-LoRA, a simple activation-guided layer selection strategy for selective Low-Rank Adaptation. We evaluate this strategy for both encoder-only and decoder-only architectures using the GLUE benchmark. Our method achieved a 20% GPUh saving with a 1% drop in GLUE score using DeBERTaV3-Base on a single-instance GPU with 50% less LoRA parameters. It also achieved 2% GPUh savings with a less than 0.15% drop in GLUE score with the Llama-3.1-8B model in Distributed Data Parallel mode with 25% fewer LoRA parameters. Our experiments and analysis show that the compute and memory requirements of LoRA adapters increase linearly with the number of selected layers. We further compare activation-guided selection against gradient-guided importance metrics and show that activation norms yield more stable and reproducible layer rankings across seeds and datasets. Overall, our results demonstrate that activation-guided layer selection is a practical and effective way to improve the efficiency of LoRA fine-tuning, making it immediately compatible with some existing PEFT techniques and distributed training frameworks. Full article

Colorectal cancer (CRC) develops through both conventional adenoma–carcinoma and serrated neoplasia pathways, yet noninvasive screening still under-detects the advanced precursor lesions that enable true cancer prevention. Stool-based screening reduces CRC mortality, but its preventive impact remains constrained by limited detection of advanced precancerous lesions (APLs), including advanced adenomas and sessile serrated lesions. Next-generation multitarget stool DNA assays (mt-sDNA; e.g., Cologuard Plus) have established high sensitivity for CRC and specificity approaching 94%, leaving improved APL detection as the principal opportunity for innovation. This review presents a consensus framework for a multi-omic stool screening paradigm that integrates host epigenetic markers (DNA methylation) with gut microbiome features using artificial intelligence (AI). Multi-omics capture complementary layers of early tumor biology: epithelial shedding and field effects reflected in host methylation signals together with luminal ecological and inflammatory changes represented by microbial features. Evidence from cross-cohort microbiome studies indicates that microbial signatures provide an additive—rather than standalone—axis of information for CRC and its precursor lesions. Because microbiome-based models are highly susceptible to batch effects arising from collection devices, extraction chemistry, sequencing platforms, and bioinformatic pipelines, practical mitigation strategies are outlined, including harmonized pre-analytics, batch-aware study design, leakage-resistant validation, and computational harmonization. A translational roadmap linking analytical validity, locked-model development, and prospective colonoscopy-verified clinical validation is proposed, aligned with TRIPOD + AI, STARD, PROBAST-AI, SPIRIT-AI, CONSORT-AI, and DECIDE-AI reporting standards. Scenario modeling using BLUE-C prevalence estimates suggests that improving APL sensitivity from approximately 43% to 55–65% at ~94% specificity could translate to detecting roughly 13–23 additional advanced precancerous lesions per 1000 individuals screened, highlighting the potential prevention impact of a multi-omic approach. This framework aims to guide developers and clinical investigators toward next-generation stool tests capable of materially improving precursor-lesion detection while maintaining clinically acceptable specificity. Full article

Background/Objectives: The popliteus muscle (PM) plays a crucial role in stabilizing the posterolateral aspect of the knee. However, its layered structure and innervation are not well understood due to its location, size, and proximity to neighboring anatomical features. This study aimed to clarify the layered morphology, intramuscular innervation, and fiber-type composition of the PM, providing anatomical insights for clinical interventions. Methods: We examined 32 lower extremities from sixteen formalin-embalmed cadavers using a multimodal approach that included gross dissection, Sihler’s staining, ultrasonography, and histochemical analysis. Results: On average, 2.8 ± 1.1 branches of the tibial nerve entered the PM, with a consistently high-density entry zone located at 56–64% of the muscle length. Sihler’s staining and ultrasonographic analyses revealed a distinct separation between the superficial and deep layers across the central tendon, each exhibiting compartmentalized intramuscular branching territories. The superficial layer was primarily composed of type IIx fibers and exhibited a larger pennation angle, while the deep layer was richer in type IIA fibers with a smaller pennation angle. These findings illustrate that the PM functions as a dual motor unit rather than a uniform structure. Conclusions: The PM exhibits a distinct compartmentalized organization, functioning as a multifunctional motor unit. The identification of specific intramuscular entry zones and the organization of muscle layers provide strong anatomical evidence for improved targeting in neuromuscular-modulating interventions. This enhances the precision, safety, and efficacy of clinical strategies aimed at addressing posterior knee stability and pathologies related to the posterolateral complex (PLC). Full article

The existing research on Android malware detection using graph neural networks (GNNs) has largely focused on architectural improvements, while input node feature representations have received less systematic attention. This study adopts a representation-centric approach to enhance function call graph (FCG)-based malware classification through interpretability-driven feature engineering. We propose a dual-level structural feature framework integrating local topological patterns with global graph-level properties. The initial feature set comprises 13 dimensions: five local degree profile (LDP) features and eight global structural features capturing community structure, execution flow, and connectivity patterns. To mitigate the curse of dimensionality, we apply an interpretability-driven selection using integrated gradients (IG), gradient-weighted class activation mapping (GradCAM), and Shapley additive explanations (SHAP), yielding an optimized seven-dimensional subset. Experiments on the MalNet-Tiny benchmark demonstrate that the proposed approach achieves 94.47 ± 0.25% accuracy with jumping knowledge GraphSAGE (JK-GraphSAGE), improving the LDP-only baseline by 0.32 percentage points while reducing feature dimensionality by 46%. The selected features exhibit consistent importance across four GNN architectures and multiple message-passing layers, demonstrating model-agnostic effectiveness. The results reveal that aggregation mechanisms critically influence feature utility, highlighting the necessity of interpretability-guided design for robust malware detection. This work provides a systematic methodology for feature engineering in graph-based security applications. Full article

Value investing remains grounded in intrinsic value estimation, margin-of-safety reasoning, and disciplined fundamental analysis, but its practical execution is increasingly constrained by the scale, heterogeneity, and velocity of modern financial information. Recent advances in artificial intelligence (AI), particularly large language models and automated information-extraction systems, create new opportunities to accelerate financial analysis; however, their outputs remain probabilistic, context-dependent, and potentially error-prone, making governance and verification essential. This article proposes an AI-assisted value investing framework that integrates automated extraction, valuation modeling, explainability, and human-in-the-loop (HITL) supervision into a unified decision-support architecture. The framework is organized into three layers: (i) a data layer for traceable extraction and normalization of structured and unstructured financial information; (ii) a modeling layer for automated key performance indicator (KPI) computation, forecasting support, and discounted cash flow (DCF) valuation; and (iii) an explainability and governance layer for traceability, verification, model-risk control, and analyst oversight. A central contribution of the paper is the operational characterization of prompt literacy as a determinant of analytical reliability, showing that structured, context-aware prompts materially affect extraction correctness, usability, and verification effort. The framework is evaluated through a case study using Rivanna AI on three large U.S. beverage firms—namely, The Coca-Cola Company, PepsiCo, and Keurig Dr Pepper—selected as a controlled, information-rich setting for comparative analysis. The results indicate that the proposed workflow can reduce end-to-end analysis time from approximately 25–40 h in a traditional manual process to approximately 8–12 h in an AI-assisted setting, including citation/source verification, unit and period reconciliation, and review of key valuation assumptions. Rather than eliminating analyst effort, AI shifts it from manual information processing toward verification, adjudication, and interpretation. Overall, the findings position AI not as an autonomous decision-maker, but as a governed reasoning accelerator whose effectiveness depends on structured human guidance, traceability, and disciplined validation. In value investing, a discipline traditionally grounded in labor-intensive fundamental analysis and disciplined intrinsic value estimation, AI introduces the potential to scale analytical coverage and accelerate evidence synthesis. However, AI systems in financial contexts are probabilistic, context-sensitive, and inherently dependent on human interaction, raising critical questions about reliability, governance, and operational integration. This article proposes a structured framework for AI-driven value investing that preserves the foundational principles of intrinsic value, margin of safety, and economic reasoning, while redesigning the analytical workflow through automation, explainability, and human-in-the-loop (HITL) supervision. The proposed architecture integrates three layers: (i) an AI-enabled data layer for traceable extraction and normalization of structured and unstructured financial information; (ii) a modeling and valuation layer combining automated KPI computation, machine learning forecasting, and discounted cash flow (DCF) valuation; and (iii) an explainability and governance layer ensuring traceability, verification, and model risk control. A central contribution of this work is the operational characterization of prompt literacy, namely the ability to formulate structured, context-aware requests to AI systems, as a critical determinant of system reliability and analytical correctness. Through a focused case study using an AI-assisted analysis platform (Rivanna AI) on three U.S. beverage firms, we provide evidence that structured prompt formulation can improve extraction consistency, reduce verification overhead, and increase workflow efficiency in a human-supervised setting. In this setting, analysis time decreased from a manual range of approximately 25–40 h to 8–12 h with AI assistance and HITL validation, while preserving traceability and decision accountability. The reported hour savings should be interpreted as conservative estimates from the initial deployment phase; additional efficiency gains are expected as operational maturity increases, driven by learning-economy effects. The findings position AI not as an autonomous decision-maker but as a probabilistic reasoning accelerator whose effectiveness depends on structured human guidance, verification discipline, and prompt-driven interaction. These results redefine the role of the financial analyst from manual data processor to reasoning architect, responsible for designing, guiding, and validating AI-assisted analytical workflows. Full article

To address the problems of severe fruit occlusion, large variations in target scale, and many small-scale goals being overlooked in the recognition of trellised watermelons under complex agricultural scenarios, this study proposes an improved RT-DETR-based detection model, termed RT-DETR-Watermelon. A context-guided (CG) module is embedded into the backbone network. A dedicated P2 detection layer is added to enhance the model’s sensitivity to small objects. A scale sequence feature fusion (SSFF) module and a triple feature encoder (TFE) module are introduced into the model to improve the model’s capability to detect targets at multiple scales. The original bounding box regression loss is replaced with MPDIoU (Multiple Path Distance Intersection over Union) loss, which accelerates model convergence and improves localization precision. Finally, the number of channels is adjusted to reduce parameter count, computational complexity, and storage size. The experimental results show that, compared with the original RT-DETR model, the proposed RT-DETR-Watermelon model increases precision, recall, and mean Average Precision (mAP@0.5) by 0.4, 1.8, and 1.0 percentage points, while reducing the number of parameters, computational cost, and model size by 53.5%, 23.5%, and 53.2%, respectively. Full article

With the global population predicted to reach 10 billion by 2050, pesticides are essential for agricultural production. However, they can introduce chemical stressors into aquatic ecosystems. Chlorpyrifos (CPF) is a widely used organophosphate insecticide that can enter aquatic environments and poses potential risks to early-life-stage fish. Because the retina is an extension of the central nervous system and vision-guided behaviors are highly sensitive to neural dysfunction, we hypothesized that CPF exposure disrupts neurobehavioral and visual function via oxidative stress and PPARα-related signaling. Zebrafish larvae were exposed to CPF (0.01, 0.1, 1, 10, and 100 μg/L) with a vehicle control (VC). During the photomotor response assay, exposure to 100 μg/L CPF reduced overall swimming activity by 48.90% and dark-period activity by 57.71%, whereas 1 μg/L CPF modestly increased total distance by 6.96% (p = 0.003) and dark-period distance by 5.40% (p = 0.011). Transcriptomic profiling highlighted nervous- and vision-related functional categories, and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment implicated pathways including gonadotropin-releasing hormone (GnRH), mitogen-activated protein kinase (MAPK), and peroxisome proliferator-activated receptor (PPAR) signaling. Targeted neurotransmitter metabolomics showed significant increases in dopamine, γ-aminobutyric acid (GABA), and acetylcholine across treatment groups, indicating broad neurotransmitter dysregulation. Consistent with these findings, neuronal fluorescence in Tg (elavl3: EGFP) larvae decreased by 12.1% and 32.5% in the 1 and 100 μg/L groups, respectively (p < 0.001), and glial fibrillary acidic protein (GFAP) immunofluorescence increased in the eye/brain/olfactory bulb at 1 μg/L (p = 0.037) and 100 μg/L (p = 0.002). Histology further showed retinal injury, with a 14.3% reduction in photoreceptor layer thickness at 100 μg/L (p = 0.034). Mechanistically, coexposure to a PPARα antagonist (GW6471) alleviated CPF-induced behavioral deficits (1.80-fold increase in dark locomotion) and reduced elevated GABA and dopamine levels by 36.8% and 47.3%, respectively. Together, these results indicate that CPF can impair neuro-visual development and that oxidative stress and PPARα-related signaling are closely associated with these effects. Full article