Platinum group of metals (PGMs) are extensively used in numerous arenas as they have discrete physical and chemical properties. Platinum, Palladium, Rhodium, Ruthenium, Iridium, and Osmium are the precise metals that have high catalytic activities, significant electrical conductivity, and high corrosion resistance. Based on the conventional technologies to recover platinum group of metals, several methodologies have been suggestively enhanced. Due to the lack of primary sources and the increased economic value of these precise metals such as Palladium, it is crucial to recover these metals from different secondary resources, like, spent automotive catalytic converters, waste electronic and electrical instruments, used batteries, various low-grade ores, etc. Compared with traditional separation processes, the bioleaching technology with varying parametric conditions has rewards in terms of significant recovery efficiency, economically beneficial, good mechanism of desorption, and renewability. This overview inaugurates by introducing the platinum group of metals (PGMs) mainly palladium with their various application towards biomedical usage. Subsequently, the assessment of the recovery of palladium using bioleaching has been discussed in this synopsis.
A new approach and catalyst system for the simultaneous heterogeneous fluid-catalytic conversion and recovery (SFCCR) of bio-aviation fuel from Jatropha seed oil (JO) was investigated. The oxalates of bimetallic nano-particles of Nickel–Molybdenum supported on activated carbon (NiMo-Ox/AC) were synthesized and used as the catalyst. The nonlinear kinetics, thermodynamics, and effects of process variables on bio-aviation fuel from Jatropha seed oil (JO) were studied. The nonlinear kinetic models of pseudo-first-order and pseudo-second-order models were investigated. The process variables studied include the SFCCR temperature, catalyst loading, and conversion time. The thermodynamics parameters such as Gibb free energy, enthalpy, and entropy were determined. The results revealed that bio-aviation fuel yield increased with an increase in temperature, catalyst loading, and conversion time. The maximum bio-aviation fuel yield of 90
The rapid growth of online communication has increased the need for efficient multimodal sarcasm detection models. However, effectively integrating textual, visual, and audio information remains a significant challenge. To tackle this issue, this paper proposes a Conditional Random Field-assisted LinkNet framework (CRF-LN-MSD) that integrates aspect-level textual representations via a Modified Fisher Score-based Aspect Term Extraction (MFS-ATE) technique with facial expression features and audio spectral characteristics. The CRF-LN architecture employs a CRF layer for structured multimodal fusion and a novel Tanh-Swish-Sigmoid (TSS) activation function for improved gradient flow. Experimental results on the M2H2 and CMU-MOSI datasets demonstrate competitive performance under the evaluated experimental conditions, with a precision of 0.978 on M2H2 and a cross-validated accuracy of 0.9787 on CMU-MOSI. These results demonstrate the effectiveness of the proposed architectural refinement for multimodal sarcasm detection within the evaluated scope.
Industry 4.0 applications require deterministic wireless communication with strict guarantees on latency, reliability and deadline compliance. IEEE TSN and 3GPP 5G URLLC are among the existing standards that can serve as a basis for achieving synergies between heterogeneous wireless networks. This article proposes a standardized middleware architecture that enables deterministic communications on multi-radio industrial networks using 5G URLLC, Wi-Fi 6E and TSN. Industry 4.0 applications where the proposed middleware architecture can be applied. The paper introduces a unified middleware architecture capable of enabling deterministic communication over multi-radio industrial networks through 5G URLLC, Wi-Fi 6E and TSN technologies. The proposed Deadline-Aware Resource Scheduling (DARS) mechanism adapts resource allocation based on the traffic characteristics and the network status. The experimental simulation indicates that DARS achieves a close to 0
The harmful use of Deepfake technique is a major risk to public safety, causing a crisis in public opinion. Despite numerous attempts to identify deepfake videos, current approaches lack generalization. These approaches often lack robustness, struggle with generalization across different types of deepfakes, and fail when the manipulations are subtle or generated by advanced models. As deepfake generation technology evolves, it consistently outpaces traditional detection techniques, necessitating more sophisticated, adaptive, and learning-based detection systems. Thus, this article proposes an effective Cascaded Neuro Fuzzy Parallel Convolutional Forward Harmonic Network (Cascaded NFPCFHNet) for deepfake recognition. At first, frame extraction is conducted. Then, Deep Neural Network (DNN) performs face detection on the extracted frames and Facial Action Unit (AU) detection is performed by AUNet. Moreover, Feature extraction is performed for each AU, capturing essential features for improving the detection accuracy. Finally, Deepfake Recognition is done using Cascaded NFPCFHNet. The Cascaded NFPCFHNet is a novel architecture formed by combining a Cascaded Neuro Fuzzy Network (CNFN) and a Parallel Convolutional Neural Network (PCNN) with Harmonic analysis. Moreover, Cascaded NFPCFHNet attained accuracy, True Positive Rate (TPR), as well as True Negative Rate (TNR) of 91.892
Asthma is a heterogeneous chronic airway disease driven by complex interactions among genetic, environmental, metabolic, and microbial factors. Conventional diagnostic and therapeutic approaches often fail to capture its molecular diversity, underscoring the need for integrative strategies. Metabolomics and metagenomics have emerged as transformative tools for dissecting asthma pathophysiology, offering complementary insights into host biochemical alterations and microbiome-mediated immune modulation. Metabolomic studies across biofluids such as urine, plasma, exhaled breath condensate, sputum, and bronchoalveolar lavage fluid have identified alterations in lipid, amino acid, and short-chain fatty acid metabolism associated with asthma severity and inflammatory endotypes. Metagenomic analyses have further revealed microbial dysbiosis in asthma, including increased airway abundance of Haemophilus, Moraxella, and Neisseria, along with reduced commensal taxa. Early-life gut microbiome changes, such as reduced diversity and increased fungal taxa like Candida and Pichia, have also been linked to increased asthma risk. Recent advances in systems biology and artificial intelligence (AI) have enabled integration of multi-omics datasets to uncover host–microbe–metabolite networks that drive asthma heterogeneity. Machine learning and network-based approaches have enabled the accurate identification of predictive biomarkers, delineation of disease endotypes, and prioritization of potential therapeutic targets. Despite these advances, the widespread implementation of AI-driven multi-omics approaches in asthma is challenged by algorithmic biases, data security and privacy concerns, and the black-box nature and limited interpretability of these models, necessitating rigorous clinical validation before their translation into clinical practice. This review provides a comprehensive overview of metabolomic and metagenomic alterations in asthma, explores their convergence through microbial–metabolic crosstalk, and highlights the potential of AI-guided systems biology approaches to advance next-generation precision respiratory medicine.
We introduce a mathematically consistent Gamma–Lindley q-distribution and develop a support-aware deep probabilistic regression framework for positive-valued responses. The proposed model is constructed within Jackson q-calculus and uses the q-gamma function to obtain an explicit normalization mechanism together with closed-form expressions for raw moments, the conditional mean, and the conditional variance. We establish positivity, exact normalization, and convergence to the classical admissible Gamma–Lindley distribution as q → 1^- . Building on these analytic results, we embed the proposed distribution into a neural conditional density estimator whose parameters depend on the covariates through constraint-preserving output maps. In particular, the model learns the conditional support endpoint explicitly, which guarantees likelihood feasibility and stabilizes training under finite q-support. Numerical experiments validate the q-gamma Jackson-integral representation, normalization and moment identities, sampling on the Jackson grid, and the classical-limit behavior. In synthetic probabilistic-regression experiments, the identifiable version of the proposed model accurately recovers the latent parameter functions while providing sharper predictive intervals and substantially improved conditional-mean estimation compared with a baseline likelihood-only model. We further benchmark the proposed estimator against Gamma, Weibull, log-normal and generalized-gamma deep distributional-regression baselines on both synthetic data and a real positive-valued dataset, where it attains competitive predictive accuracy together with well-covered predictive intervals. These results show that the Gamma–Lindley q-distribution provides a theoretically grounded and practically effective foundation for deep probabilistic regression of positive data.
Tomato leaf diseases threaten global food security by reducing crop yield and quality. Centralized deep-learning detectors require farms to share raw images, which raises privacy, bandwidth, and data-sovereignty concerns. We present FedPer-KD, a privacy-aware federated framework that combines personalization with client-side knowledge distillation: each client keeps its raw images and personalized classifier head locally while the shared MobileNet-V2 feature extractor is aggregated, and a local EfficientNet-B0 teacher supervises the student through softened logits. We evaluate on the PlantVillage tomato leaf dataset (18,345 images, ten classes) using simulated federated clients (N = 2, 4, 6) trained for 10, 20, and 30 rounds with five random seeds. FedPer-KD reaches 99.34
This study presents a mathematical model describing the interaction between human immunodeficiency virus (HIV) and CD4+ T cells within a predator–prey framework, where HIV acts as the predator and CD4+ T cells serve as the prey. Unlike previous simplified models that treat infected cell death as a constant process, this work introduces an explicit density-dependent cytotoxic immune response mechanism that eliminates infected CD4+ T cells in proportion to both the infected cell and viral populations. The model comprises three ordinary differential equations representing uninfected CD4+ T cells, infected CD4+ T cells, and free virus particles. Mathematical analysis reveals two equilibrium points: a virus-free equilibrium and an endemic equilibrium where all three populations coexist. Stability analysis using linearization and Routh-Hurwitz criteria demonstrates that the virus-free equilibrium is locally asymptotically stable when the basic reproduction number, R0 falls below unity, while the endemic equilibrium achieves stability under sufficient conditions involving the immune response strength. Numerical simulations employing parameter values informed by recent immunological studies confirm that robust cytotoxic immune responses promote stable chronic infection with reduced viral burden. Global stability analysis via Lyapunov functions and bifurcation analysis further validate these findings. The results highlight the critical role of adaptive immunity in determining HIV persistence and provide a theoretical foundation for understanding immune control mechanisms in untreated HIV infection.
During vehicle operation, noise signals present challenges such as a broad frequency range, strong non-stationarity, and the limited effectiveness of traditional passive noise reduction methods in the low-frequency band. To address these issues, this study proposes an active noise control model that integrates the least mean square algorithm with support vector machines. The noise control architecture incorporates a least mean squares branch for real-time suppression of linear noise components, while a support vector machine branch models and compensates for nonlinear residuals and frequency drift. Outputs from both branches are coupled via a hybrid coefficient, forming a closed-loop control mechanism integrated with secondary channel modeling. Experimental results demonstrated that the proposed model achieved significantly faster convergence speed and reduced steady-state error compared to the standard FxLMS system across both datasets, with in-band noise reduction increasing by approximately 2.0–2.5 dB. Furthermore, it achieved faster stabilization control while maintaining an average computational delay of around 3.3 ms. Under frequency drift conditions, the method exhibited markedly shorter error recovery times during step and ramp transitions than the benchmark algorithm, indicating superior rapid tracking capability and control stability in complex non-stationary environments. In summary, this research not only effectively enhances noise reduction performance and nonlinear adaptability under complex automotive conditions but also provides an efficient and feasible technical approach for low-frequency noise mitigation in intelligent driving environments.
Abstract This study explores entropy generation in the magnetohydrodynamic flow of a tetra hybrid nanofluid. ( $${\text{Al}}_{2} {\text{O}}_{3} - {\text{Cu}} - {\text{SiO}}_{2} - {\text{TiO}}_{2} /{\text{H}}_{2} {\text{O}}$$ ) over a permeable surface embedded in a Darcy–Forchheimer porous medium, motivated by the need to enhance thermal system efficiency and minimise energy losses. The model incorporates key physical effects, including multiple dissipation mechanisms (Ohmic, viscous, Darcy, and Forchheimer), thermal radiation, internal heat generation, activation energy, and chemical reaction. The transformed nonlinear equations are solved numerically using the bvp4c technique. Results reveal that thermal radiation and heat source significantly elevate temperature and entropy generation, indicating increased irreversibility, while magnetic field and porous resistance strongly suppress fluid motion. Activation energy enhances species concentration, whereas a higher Schmidt number reduces mass diffusion. The Bejan number decreases under stronger magnetic and porous effects but increases with radiation and inertial resistance. The combined influence of these parameters provides deeper insight into controlling heat and mass transfer in advanced nanofluid systems. The findings are particularly relevant for applications in energy systems, cooling technologies, and thermal management, where minimising entropy generation is crucial for improving performance and efficiency.
Additive manufacturing (AM) technology has evolved considerably and its major impact on the substitution of AM products over conventional parts. Fused deposition modeling (FDM) process falls under the classifications AM technique and predominantly produces thermoplastic-based products. The existing studies have explored the performance of Carbon fiber reinforced Polylactide, the understanding of correlation between the structure and property needs to be much focused which motivates to perform this study. In the present work, Carbon fiber reinforced Polylactide was used as a filler to fabricate the tensile specimen by optimize the printing conditions includes layer thickness, infill orientation, infill density, and printing strategy. From the results it has been observed that the sample printed with 0.3 mm layer height, 45° infill orientation, 30
To address the problems of perceptual redundancy estimation in digital image watermarking, which are poor adaptability to fixed-size inputs and insufficient coordination between local and global visual features, a perceptual redundancy estimation based on perceptually sensitive size optimization is proposed. First, BI and block aggregation are combined to achieve size adaptation of visual saliency maps. A multi-scale saliency map fusion mechanism is then used to combine local details with global structural information. An adaptive weighting strategy based on edge complexity is introduced to dynamically modulate feature contributions. Finally, the optimized model is integrated into an extended transform dithering modulation watermarking framework to achieve perceptually adaptive embedding. The proposed method is validated on the USC-SIPI and BSDS300 datasets. Experimental results show that the proposed method improves structural similarity and peak signal-to-noise ratio by 5.38
To reduce the interaction noise of the cylinder-airfoil configuration, longitudinal grooves are applied to the upstream cylinder. In this work, the noise and flow characteristics are numerically investigated, with a focus on three key factors: groove profile, groove size ratio, and groove number. It is found that when the groove profile is circular, the groove size ratio is 0.01, and the groove number is 60, the far-field noise is reduced by approximately 14 dB—the maximum reduction observed. Shallow grooves exhibit a better noise reduction effect than deep ones. The investigation of transient flow fields shows that shedding vortices have a significant impact on far-field noise levels. The groove structure, especially the circular groove profile, can break down large-scale vortices into smaller ones, which contributes to noise reduction. This investigation provides new insights and methods for noise control in wind turbines and landing gear systems.
Segmentation of medical images is a fundamental step towards quantification and clinical decision support. Many approaches have been proposed and successfully demonstrated on benchmark datasets. However, reliable application of current approaches in real clinical practice remains challenging. Medical image segmentation is not a single task but rather a group of different tasks with different input and output characteristics, which call for dedicated solutions. Task difficulty and failure modes vary with anatomical priors, boundary ambiguity, and clinically meaningful error types. Therefore, rather than providing only a method summary, this review presents a task overview of medical image segmentation research from 2015 to 2026. We organize the literature according to clinical task properties and examine how different methodological choices influence benchmark results and real-world applicability. We categorize existing work into major task families, including organ segmentation, lesion and tumor segmentation, vascular segmentation, histopathology and cellular segmentation, and multi-task and cross-domain settings. For each category, we summarize representative modeling strategies and identify common challenges specific to each task. We further review widely used public datasets and evaluation metrics. In particular, we analyze how overlap, boundary, topology, and instance-level metrics capture different clinical priorities. Finally, we discuss challenges such as annotation burden, inter-institution variability, and clinical reliability. We then discuss emerging research directions, including foundation models, multi-modal and multi-task learning, and privacy-preserving federated collaboration, which may contribute to improved public health and clinical decision-making.
Conventional embankment dam safety assessments under reservoir drawdown conditions are typically based on deterministic finite element and limit equilibrium analyses. That use single representative values of hydraulic and shear strength parameters. Such approaches often neglect the inherent uncertainty and spatial variability of geomaterials leading to potential inaccuracies in predicting seepage behavior, slope stability and failure risk. Although probabilistic methods can better represent these uncertainties, their computational complexity limits their application in routine dam safety evaluations. Furthermore, the integration of physical based numerical modeling with machine learning for probabilistic seepage failure assessment remains insufficiently explored. To address these limitations, this study proposes a hybrid deterministic probabilistic framework that integrates transient seepage and stability analyses with probabilistic risk assessment and explainable machine learning. The framework was applied to the Ribb Embankment Dam in Ethiopia under steady state, rapid drawdown (RDD) and slow drawdown (SDD) conditions. Transient hydraulic and stability responses were simulated using SEEP/W and SLOPE/W based on Richards’ equation and the Morgenstern Price method. The generated hydraulic and stability indicators were then incorporated into a probabilistic failure model and used to train Random Forest (RF), XGBoost and Support Vector Machine (SVM) models. Probabilistic analysis indicated that 85.75
The use of artificial intelligence (AI) and machine learning (ML) in diagnosing and predicting lung cancer prognosis has changed the way images are interpreted, risks predicted, and decisions made for clinical purposes. Despite increased research in this area, there is little to no bibliometric analysis evaluating the global state of affairs. The purpose of this study was to examine this research area. In this study, a literature search was conducted by using the Scopus database, with emphasis on the role of AI and ML in the diagnosis and prognosis of lung cancer. Collaboration maps were analyzed using VOSviewer 1.6.20 and Biblioshiny. There was a significant growth in global scientific production with an exponential growth in number of publications during 2020–2024 period. The United States, China, and India became the major contributors to scientific production, and analysis of collaboration networks indicated the existence of effective international collaborations. Through citation analysis, the most impactful sources, authors, and institutions were found. Through keyword and thematic evolution analysis, the shift from computer-aided diagnosis and pulmonary nodule detection to radiomics, deep learning, convolutional neural networks, explainable AI, and precision medicine was identified. This paper is an exhaustive bibliometric analysis of the literature on the use of artificial intelligence and machine learning for diagnosing and prognosticating lung cancer. By identifying publication patterns, major contributors, collaboration networks, and research hot spots, this study has provided important insights into the field’s development and the future of research in this area. Bibliometric analysis mapped global AI and ML research trends in lung cancer. Scientific output increased exponentially, with rapid growth after 2021. Identified leading authors, institutions, journals, and influential studies. Highlighted deep learning and radiomics as a future direction for precision oncology.
Interdisciplinary digital media platforms receive concurrent image, speech, and text interaction tasks whose semantic urgency, computational demand, and service deadlines vary over time. This study formulates the problem as online scheduling of multimodal interaction tasks across heterogeneous GPU, CPU, and edge-computing nodes under queue-capacity, node-compatibility, and deadline constraints. A cross-modal attention module is used to generate a unified task descriptor and estimate the semantic consistency of the available modalities. The resulting semantic information is introduced into task pre-ranking and system-state construction rather than being treated as an independent recognition output. A tabular Q-learning scheduler then selects among ten fixed scheduling-rule and resource-node combinations, thereby avoiding an action space that changes with the number of pending tasks and available nodes. The experimental benchmark is constructed from public image, speech, and text datasets through an explicit intent-matching protocol, while a constraint-optimization oracle provides reference decisions for evaluating scheduling accuracy. Under the reported high-concurrency setting, the proposed method obtains an average queue-to-dispatch latency of 192 ms and an oracle-consistent scheduling accuracy of 87.2
Optical components play a pivotal role in precision industries; however, conventional polishing methods often struggle to achieve nano-level finishing. The increasing demand for nano-scale precision in optical components has led to the exploration of novel processes such as Magnetic Float Polishing (MFP). This paper addresses the research gap concerning the application of MFP for polishing plano-optical windows (POW) and reveals the complex dynamics governing MFP. This innovative method eliminates workpiece clamping forces and using passive magnetic levitation as well as control mechanisms such as hydrodynamic lift force to achieve nanoscale roughness. Parameters such as fluid flow speed, viscosity, and density of the magnetic abrasive fluid (influenced by magnetic particle volume fraction), as well as hydrodynamic forces, have been controlled during the process on the POW. Finite Element Analysis simulations illustrated that impeller rotation speed and magnetic particle volume fraction are key parameters influencing hydrodynamic forces. The vertical component of the hydrodynamic force, known as the normal force, is also affected by the distance between the workpiece and impeller, while the passive magnetic levitation force is influenced by the distance between the workpiece and the vessel bottom. It was discovered that the magnitude of the normal forces significantly affects surface roughness; to minimize surface roughness, the normal hydrodynamic forces must be optimized within a specific range. This study introduce the hydrodynamic lift forces of the abrasive fluid as a predictive mechanism to optimize the normal forces, achieving minimum surface roughness around 26 nm Ra.
This study aims to investigate the dynamic mechanical properties and energy absorption evolution of fractured red sandstone under natural conditions and acidic environments with varying pH levels. Impact loading tests were conducted on red sandstone specimens with different fracture angles (0°, 30°, 45°, 60°, and 90°) using a Split Hopkinson Pressure Bar (SHPB) system. Based on stress wave theory and the principle of energy conservation, the dynamic response characteristics and energy dissipation mechanisms of the specimens were thoroughly analyzed. The results indicate that the coupling effect of fractures and acid corrosion significantly degrades the strength characteristics of red sandstone. With an increase in the fracture angle, the dynamic compressive strength of the specimens exhibits an upward trend, while the acidic environment further exacerbates the deterioration of strength. Regarding energy evolution, an increase in the fracture angle leads to a decrease in the reflected energy ratio and an increase in the transmitted energy ratio. Furthermore, acid corrosion intensifies energy dissipation, thereby weakening the mechanical properties of the sandstone. The findings of this study can provide a scientific basis for evaluating the stability of underground engineering structures in complex environments, preventing geological hazards, and improving the fundamental theories of rock dynamics.
