Browse Articles
Discover research articles across all indexed journals
Digital restoration and feature recognition of a Qing-Dynasty vernacular dwelling based on multimodal data fusion
Abstract This study addresses the urgent challenge of digitally preserving severely damaged vernacular architecture that lacks complete historical documentation. Taking the Dan Tao’s Former Residence, a Qing Dynasty dwelling in the Jingchu region, as a case study, we propose a reproducible multimodal framework for reverse restoration. The approach integrates SLAM-based laser scanning, UAV photogrammetry, historical documentary evidence, analogy-driven HBIM construction, and knowledge-graph visualization. By bridging the semantic gaps between material remains, images, and textual records, the workflow enables high-fidelity digital modeling of complex components while decoding cultural features at multiple scales. Results demonstrate that the framework overcomes limitations of insufficient point-cloud density and missing documentation, achieving accurate 3D restoration of degraded structures and establishing a scalable cultural feature recognition system for Jingchu vernacular architecture. This research provides both methodological innovation and practical tools for conservation, offering a transferable paradigm for safeguarding endangered architectural heritage worldwide.
Automated DFU detection through GA-selected CNN ensemble with Grad-CAM interpretability
Pregnancy outcomes of obstetrician gynecologist mothers: a retrospective matched cohort study
Multi-heating-rate pyrolysis study for modeling hydrocarbon production from in-situ conversion of Chang 7 shale
Psychological stress and eating behaviors as interacting influences on patient–reported disease activity in Italian patients with inflammatory bowel disease
Multi-omics analysis reveals the effects of prenatal nutrition on carcass-related tissues in beef cattle
Abstract This study evaluated the long-term metabolic effects of prenatal nutrition in Nellore bulls. Pregnant cows ( n = 126) received mineral supplementation only (NP), protein–energy supplementation during the last trimester (PP), or supplementation throughout pregnancy (FP). At slaughter, longissimus (muscle and meat) and subcutaneous fat samples from the offspring were collected for transcriptomics and metabolomics analyses. Data were reduced using Weighted Gene Co-expression Network Analysis, followed by functional enrichment, and then integrated via Spearman’s correlations and holistic pathway analysis. Distinct molecular patterns emerged across prenatal nutrition treatments, although all groups influenced energy metabolism and cellular processes. The NP group was strongly associated with protein and lipid metabolism, highlighted by PPAR and sphingolipid signaling pathways, and key hub components including CNOT4 and tryptophan. In contrast, PP and FP groups were more closely linked to immune function, stress resilience, with enrichment of NF-kB signaling, cortisol synthesis, and hub components including TIE1 , YWHAZ , carnitine, and glutaconylcarnitine. Shared transcriptome–metabolome modules between groups displayed inverse correlations, suggesting potential antagonistic effects driven by maternal diet. Overall, these results indicate that prenatal nutrition shapes key metabolic processes in muscle, meat, and fat, offering insights to enhance meat quality and production through maternal feeding strategies.
Antifungal activity and mycotoxin-inhibiting potential of amphisin and rhamnolipids from Pseudomonas strains
Abstract Biosurfactants, due to their amphiphilic properties, are widely used across a broad spectrum of industries. In this paper, we describe experiments conducted to determine the antagonistic activity of amphisin and rhamnolipids against fungal plant pathogens. The bacterial strains used for biosurfactants production were Pseudomonas fluorescens DSS73 and Pseudomonas aeruginosa #112. Both amphisin and rhamnolipids were produced using waste raw materials from the food industry. The biosurfactants exhibited significant antifungal activity and the ability to reduce sterigmatocystin production. The most significant inhibition was observed during the initial days of the experiments, followed by a gradual decrease in biosurfactants activity. However, a long-term effect was obtained against Aspergillus amoenus DSM 1943, Penicillium nordicum DSM 12639, and Monilinia fructigena IOR 2138. Additionally, sterigmatocystin production in cultures of A. amoenus DSM 1943 and Aspergillus quadrilineatus DSM 820 was reduced by 89–99%, with the most significant results observed for rhamnolipids. Therefore, as demonstrated in this research, Pseudomonas -derived biosurfactants show great potential in plant cultivation by protecting against fungal invasion. Amphisin and rhamnolipids can be successfully applied as eco-friendly solutions to limit sterigmatocystin accumulation during crop storage.
Experimental investigation and AI-based prediction of engine performance and emissions using CeO2-enhanced biodiesel blends
Abstract The global depletion of fossil fuel resources and growing environmental concerns have driven the pursuit of sustainable alternative energy sources, with biodiesel emerging as a promising option. In the present work, experimental and AI prediction studies examine the combustion, performance, and emissions of a B20-blended fuel engine. The biodiesel was derived from used temple oil, and CeO 2 nanoparticles (NPs) were used as additives. The experiments conducted at a constant engine speed with various loads reveal that CeO 2 NPs at a 100-ppm concentration significantly improve brake thermal efficiency (BTE). The specific fuel consumption decreases with increasing CeO 2 NPs concentration, indicating enhanced fuel efficiency. The higher cylinder pressure and net heat release (NHR) further highlight the improved combustion characteristics. The emission analysis reveals a notable decline in hydrocarbons, carbon monoxide, and nitrogen oxides levels, with CeO 2 NPs at a 100-ppm concentration achieving the lowest emissions due to superior fuel atomization, improved combustion efficiency, and enhanced thermal management facilitated by the presence of CeO 2 NPs. The B20 blend with CeO 2 NPs at a 100-ppm concentration demonstrated superior performance metrics at maximum load, as evidenced by a 1.57% increase in BTE, a 2.83% rise in cylinder pressure, and a 5.95% increase in NHR compared to conventional diesel. A significant reduction in pollutant emissions, such as carbon monoxide, decreased by 66.67%, hydrocarbons by 13.51%, and nitrogen oxides by 6.04% relative to the B20 blend. The validation of experimental data using a random forest regressor model demonstrates strong predictive accuracy, characterized by low mean squared errors and high R 2 scores across various performance metrics. Overall, the findings of this work confirm that the B20 blend with CeO 2 NPs is a sustainable and efficient replacement for traditional diesel fuel.
Multi-scale analysis of stress-corrosion failure and anti-corrosion strategies for anchor bolts in Tashan coal mine
Thermal insulation of water-based acrylic coatings reinforced with APTES-functionalized silica fume nanoparticles
Splenectomy reduces shear stress and inflammation in liver endothelial cells during regeneration after partial hepatectomy in mice
Abstract Clinical and experimental findings indicate that splenectomy stimulates liver regeneration. The mechanisms of this effect are uncertain. This study aimed to assess the effect of splenectomy on the state of sinusoidal endothelial cells during liver regeneration after 70% resection in mouse model. Two series of experiments were performed. In the first series, splenectomy was performed on sexually mature male C57BL/6 mice, followed by 70% liver resection 7 days later. In the second series, only 70% liver resection was performed. The animals were withdrawn from the experiment 1, 3, or 7 days post-resection. The measurements encompassed the liver mass recovery dynamics; the AST, ALT, and albumin levels; and the proliferation activity of hepatocytes, as well as gene expression profiles, apoptosis rates, and proliferation activity of liver sinusoidal endothelial cells. Animals that underwent splenectomy before liver resection showed a decrease in the number of endothelial cells positive for VCAM-1 on days 1 and 7 post-resection and VE-cadherin-positive (CD144 +) cells on day 7. Liver resection caused an increase in the relative counts of dying (Annexin-PI +) and Ki67 + endothelial cells in the remnant. The apoptotic endothelial cell counts were significantly higher in the group with preserved spleen on day 1 post-resection, while the number of Ki67 + endotheliocytes was higher in splenectomized animals on day 1 and day 3 post-resection. Splenectomy exerts significant effect on the state of endothelial cells of the sinusoidal capillaries of the regenerating liver. The gene expression profiling studied by microarray technology using Clariom™ S Assay (mouse) revealed specific enrichment of the following signaling pathways in the endothelial cells of regenerating liver in splenectomized animals: PI3K-Akt-mTOR signaling pathway, Focal adhesion pathway, Chemokine signaling pathway. In this case, most of the differentially expressed genes were repressed. The decreased proinflammatory activation of sinusoidal endothelium results in decreased immigration of lymphocytes, notably CD3 + и NK1.1 lymphocytes known to inhibit the regeneration. The results indicate that prior splenectomy mitigates the manifestation of shear stress in endothelial cells of the liver and reduces their pro-inflammatory activation, resulting in lower rates of migration of regeneration-inhibiting lymphocytes to the regenerating liver.
High-performance hydrogen detection of nanoarrays based on plasmonic enhancement mechanism
Reduced cloud cover errors in a hybrid AI-climate model through equation discovery and automatic tuning
Abstract Cloud-related parameterizations remain a leading source of uncertainty in climate projections. Although machine learning holds promise for Earth system models (ESMs), many data-driven parameterizations lack interpretability, physical consistency, and smooth integration into ESMs. Here, a two-step method is presented to improve a climate model with data-driven parameterizations. First, we incorporate a physically consistent cloud cover parameterization—derived from storm-resolving simulations via symbolic regression, preserving interpretability while enhancing accuracy—into the ICON global atmospheric model. Second, we apply the gradient-free Nelder–Mead optimizer to automatically recalibrate the hybrid model against Earth observations, tuning in nested stages (2-, 7-, 30- and 365-day runs) to ensure stability and tractability. The tuned hybrid model substantially reduces long-standing biases in cloud cover—particularly over the Southern Ocean (by 75%) and subtropical stratocumulus regions (by 44%)—and remains robust under +4K surface warming. These results demonstrate that interpretable machine-learned parameterizations, paired with practical tuning, can efficiently and transparently strengthen ESM fidelity.
Interaction-induced magnetotransport in a 2D Dirac–Heavy hole hybrid band system
The application of ABC-VED with multi-criteria analysis for drug inventory management
Optical analysis of 3D-printed terahertz waveplates from common thermoplastics
CuFe2O4 nanoparticles via thermal decomposition as recyclable magnetic catalysts for perimidine synthesis
Characterization of broad host range bacteriophages vKpIN31 and vKpIN32 against hospital-acquired Klebsiella pneumoniae in Dakar, Senegal
An explainable hybrid framework for early detection of cardiovascular diseases using Categorical Boosting and Bees algorithm
Abstract Cardiovascular disease (CVD) remains one of the leading causes of death worldwide, claiming millions of lives each year. The early detection of CVD enables healthcare professionals to make informed decisions about the patient’s health. Machine learning (ML)- based frameworks have been extremely popular in predicting diseases. However, results generated from traditional ML models are “black-box,” lacking transparency and interpretability. The objective of the present study is to develop an ML framework that detects CVD with promising accuracy and, further, provide interpretability to the generated outcomes to ensure targeted therapies. The Framingham, Massachusetts CVD dataset, which is publicly available from the Kaggle Repository, is used in this study. As part of the data pre-processing, the Random Oversampling (RO) technique is applied to overcome the data imbalance problem, followed by Pearson Correlation analysis to understand the correlation between attributes. Then, the Min–Max scaling technique is used for data normalization. The pre-processed data is fed into a hybrid ML framework incorporating the Categorical Boosting (CatBoost) and BEEs algorithms to achieve optimized CVD prediction results. The proposed Hybrid model yielded 98.04% accuracy, a Precision of 97.09%, a Recall of 98.96%, an F1-score of 98.02%, and a Specificity of 97.16%, with a total execution time of 26.6580 s. The proposed model outperformed contemporary state-of-the-art algorithms, considering most evaluation metrics. Additionally, Explainable Artificial Intelligence (XAI) techniques, such as LIME and SHAP, are implemented to identify the contribution of the most significant attributes towards the occurrence of CVD, offering valuable insights into the detection of the disease and enabling healthcare providers to make accurate and timely treatment decisions.