Search Results (122)

Open-source and commercial fifth-generation (5G) deployments are difficult to compare because they are built for different goals and reported under different conditions, which slows down validation and technology transfer from research to practice. This study explores the deployment and evaluation of two 5G Standalone (SA) disaggregated Radio Access Network (RAN) systems, using open-source research RAN, commercial RAN, and Software-Defined Radio (SDR) hardware. The first testbed is a SDR-based prototype, containing a Universal Software Radio Peripheral (USRP) B210 device, using Software Radio System RAN (srsRAN) as the RAN. The commercial-based testbed contains a Benetel RAN550 Radio Unit (RU), connected via an optical fiber to a Commercial Off-the-Shelf (COTS) server acting as the Distributed Unit (DU) and Centralized Unit (CU) using the Accelleran virtualized Baseband Unit (vBBU) platform. The Core Network (CN) is implemented using the open-source Open5GS in both testbeds. To evaluate the network’s functionality, throughput and latency are tracked using a Motorola Edge 50 Pro mobile terminal. The experimental results are analyzed and compared with representative performance metrics reported in the literature to place the measurements in a broader research context. This study further assesses trade-offs related to cost, portability, and scalability by comparing SDR-based research prototypes with commercial deployments. Full article

Manufacturing companies face significant challenges in leveraging artificial intelligence for equipment management due to high infrastructure costs and limited availability of labeled data for failures. While most manufacturing AI applications focus on structured sensor data, vast amounts of unstructured textual information containing valuable maintenance knowledge remain underutilized. This study presents a practical generative AI-based framework for structured information extraction that automatically converts unstructured equipment maintenance texts into predefined semantic fields to support predictive maintenance in manufacturing environments. We adopted and evaluated three representative generative models—Bidirectional and Auto-Regressive Transformers (BART) with KoBART, Text-to-Text Transfer Transformer (T5) with pko-t5-base, and the large language model Qwen—to generate structured outputs by extracting three predefined fields: failed components, failure types, and corrective actions. The framework enables the structuring of equipment management text data from Manufacturing Execution Systems (MES) to build predictive maintenance support systems. We validated the approach using a large-scale MES dataset consisting of 29,736 equipment maintenance records from a major automotive parts manufacturer, from which curated subsets were used for model training and evaluation. Our methodology employs Generative Pre-trained Transformer 4 (GPT-4) for initial dataset construction, followed by domain expert validation to ensure data quality. The trained models achieved promising performance when evaluated using extraction-aligned metrics, including exact match (EM) and token-level precision, recall, and F1-score, which directly assess field-level extraction correctness. ROUGE scores are additionally reported as a supplementary indicator of lexical overlap. Among the evaluated models, Qwen consistently outperformed BART and T5 across all extracted fields. The structured outputs are further processed through domain-specific dictionaries and regular expressions to create a comprehensive analytical database supporting predictive maintenance strategies. We implemented a web-based analytics platform enabling time-series analysis, correlation analysis, frequency analysis, and anomaly detection for equipment maintenance optimization. The proposed system converts tacit knowledge embedded in maintenance texts into explicit, actionable insights without requiring additional sensor installations or infrastructure investments. This research contributes to the manufacturing AI field by demonstrating a comprehensive application of generative language models to equipment maintenance text analysis, providing a cost-effective approach for digital transformation in manufacturing environments. The framework’s scalability and cloud-based deployment model present significant opportunities for widespread adoption in the manufacturing sector, supporting the transition from reactive to predictive maintenance strategies. Full article

To evaluate cross-platform measurement consistency and diagnostic threshold requirements in shear wave elastography (SWE), this study presents a robotically controlled, phantom-based validation framework to quantify and interpret inter-vendor variability that limits clinical standardisation. A custom-fabricated polyvinyl chloride-graphite phantom containing eight spherical inclusions (15–25 mm diameter, 25–95 mm depth, 23.53–259.58 kPa stiffness), representing breast tissue mechanical properties, was evaluated using Samsung HS50 and Aixplorer ultrasound systems. Robotic automation standardised probe positioning and contact, eliminating operator-dependent variability and enabling direct, system-level comparison. Cross-platform reproducibility, accuracy against mechanically validated ground truth, and diagnostic threshold performance were assessed across 80 measurements. Both systems demonstrated excellent intra-machine reproducibility (coefficient of variation: Samsung 0.42%, Aixplorer 0.46%) with strong inter-machine correlation (r = 0.9951, p < 0.0001). However, systematic bias of 7.05 kPa and 95% limits of agreement spanning 38.7 kPa revealed substantial cross-platform measurement differences. All phantom inclusions (8/8) measured below their assigned ground truth stiffness on both systems, with systematic underestimation ranging from 0.33 kPa to 109.57 kPa, indicating parameter-dependent measurement distortion linked to inclusion size, depth, and stiffness. Dynamic range compression was observed (Samsung: 68.7%, Aixplorer: 59.1% of true phantom range), providing a mechanistic explanation for diagnostic threshold instability. This study contributes an interpretable validation methodology that links SWE measurement bias to physical lesion properties and imaging system characteristics, rather than relying on correlation alone. Despite strong reproducibility, the observed system-dependent bias demonstrates that SWE measurements are not directly transferable across ultrasound platforms, and system-specific collaboration is required to ensure reliable clinical interpretation. Full article

BNCT (Boron Neutron Capture Therapy) is a binary radiotherapeutic modality in which high LET (Linear Energy Transfer) particles are generated from ¹⁰B(n,α)⁷Li reaction, ideally within boron-loaded tumour cells, so the therapeutic outcome depends critically on the pharmacokinetics and biodistribution of boron carriers. In this review, boron-containing agents for BNCT, with a focus on ADMET (absorption, distribution, metabolism, excretion and toxicity) and model-informed design, were examined. Low-MW (low-molecular-weight) compounds, peptide conjugates, polymeric and nanostructured platforms and cell-based vectors were surveyed and how physicochemical properties, transporter engagement and nano–bio interactions govern tumour uptake, subcellular localisation and normal tissue exposure were discussed. A shift from maximising boron content towards optimising exposure profiles using PET (Positron Emission Tomography), PBK (physiologically based pharmacokinetic) modelling and in silico ADMET tools to define irradiation windows was also discussed. Classical agents such as BPA (Boronophenylalanine) and BSH (Sodium Borocaptate) are contrasted with newer polymeric and metallacarborane-based carriers, with attention to brain penetration, endosomal escape, linker stability, biodegradation and elimination routes, as well as platform-specific toxicities. Incontestably, further progress in BNCT will highly depend on integrating imaging-derived kinetics with PBPK-informed dose planning and engineering subcellularly precise yet degradable carriers, and that ADMET-guided design and spatiotemporal coordination are central to achieving reproducible clinical benefit from BNCT’s spatial selectivity. Full article

Early diagnosis of Alzheimer’s disease represents a critical clinical challenge, and the high-sensitive biomarkers measurement holds great potential for enabling early identification and intervention. This study proposes an electrochemical immunosensing strategy based on inkjet printing for the quantitative detection of Tau-441. Conductive patterns were formed by inkjet printing, followed by surface functionalization with gold nanoparticles to immobilize highly specific anti-Tau-441. This process created a stable and high affinity immunorecognition interface that enhances electron transfer and signal amplification. Furthermore, we developed and integrated a low-power portable detection platform to achieve a rapid detection process encompassing sample loading, signal acquisition, and on-device readout. The method shows a linear response from 50 fg/mL to 10 ng/mL and a limit of detection of 16 fg/mL (S/N = 3), with high specificity and good reproducibility. By combining scalable inkjet fabrication with a self-contained handheld reader, this method shortens the path from sample to result and offers a practical route for on-site screening and early intervention in Alzheimer’s disease. Full article

Sulfated nanocellulose (SCNF) and reduced graphene oxide (rGO) films were fabricated through environmentally friendly methods to develop an effective platform for electrochemical applications. The hybrid materials were extensively characterized by FTIR, XRD, Raman spectroscopy, TGA, SEM, cyclic voltammetry (CV), and electrochemical impedance spectroscopy (EIS). Results showed that incorporating rGO into the SCNF matrix significantly improved the electrical conductivity and structural robustness of the films. FTIR confirmed interactions between sulfate groups on cellulose and residual oxygen-containing groups on rGO, while XRD and Raman analyses indicated reduced crystallinity and increased structural disorder, supporting the successful integration of both phases. XPS further demonstrated that SCNF and rGO form chemical bonds rather than simply mixing, with both components remaining active at the surface—evidence of strong interfacial interactions that contribute to enhanced stability and efficient charge transfer. The 1:5 (rGO:SCNF) composition showed the best electrochemical performance, exhibiting minimal charge-transfer resistance and improved hydrazine oxidation, as reflected by a shift of the anodic peak potential toward lower values. Additionally, functionalization with cobalt porphyrin significantly boosted catalytic activity. Overall, the SCNF:rGO films offer a sustainable and scalable platform for electrochemical sensing and energy-conversion applications, demonstrating excellent adaptability and functional performance. Full article

Background: Cavariella salicicola (Hemiptera: Aphidinae) is a pest on Salix spp. and various Umbelliferae (Apiaceae) vegetables. However, the taxonomic status and phylogenetic relationship of the genus Cavariella within Aphidinae remain controversial due to the small body size and easily confused external morphology. Methods: The complete mitochondrial genome of C. salicicola collected from Oenanthe javanica was sequenced using the Illumina platform and compared with C. theobaldi. The codon usage bias of two Cavariella aphids was assessed through Enc plot, PR2 plot, and neutrality plot analyses. Furthermore, phylogenetic trees were constructed based on both Maximum Likelihood and Bayesian Inference analysis. Results: The C. salicicola mitochondrial genome comprises 15,720 bp and represents a typical circular DNA molecule with a high AT content of 83.8%. It contains the standard 37 genes, including 2 ribosomal RNAs (rRNAs), 13 protein-coding genes (PCGs), 22 transfer RNAs (tRNAs), and 2 long non-coding regions (control and repeat regions). Varying degrees of codon usage bias were found across different PCGs, and the bias was predominantly influenced by natural selection rather than mutational pressure. The ratio of nonsynonymous to synonymous substitutions (Ka/Ks) indicated that all PCGs in C. salicicola, as well as most other Aphidinae species, are under strong purifying selection. The phylogenetic analysis based on Maximum Likelihood and Bayesian Inference both strongly supported the monophyly of Aphidinae, Macrosiphini, and Aphidini. Crucially, the monophyletic genus Cavariella was resolved as a sister group to all other sampled species within the tribe Macrosiphini. Conclusions: This study provides new molecular data to support the sister relationship of the genus Cavariella to other Macrosiphini aphids. This study will enhance our understanding of phylogenetic relationships within the subfamily Aphidinae. Full article

Embryo genomic selection (EGS) is a contemporary breeding strategy that combines genomic selection (GS) methodology with embryo biotechnology. By conducting genotyping and genomic prediction at the pre-implantation stage, embryos with superior breeding value can be identified for transfer, markedly increasing breeding efficiency while reducing the uncertainty and temporal expenditure associated with conventional GS. This study establishes a reliable embryo biopsy-based GS pipeline for sheep, incorporating optimized whole-genome amplification and microcell genotyping techniques. We developed a high-efficiency in vitro sheep embryo production platform compatible with embryo biopsy. Systematic comparison of Multiple Displacement Amplification (MDA) and Multiple Annealing and Looping Based Amplification Cycles (MALBAC) whole-genome amplification systems yielded high-quality genotypes from biopsy samples of embryos containing as few as 10 cells. Imputation using 10× whole-genome sequencing data significantly increased both genotype call rates and accuracy. High concordance was observed between embryo and lamb genotypes, and genomic estimated breeding values (GEBVs) for key growth traits exhibited strong correlations (R²: 0.91–0.98). This system enables accurate preimplantation genomic evaluation and provides an efficient strategy to accelerate genetic improvement in sheep breeding programs. Full article

Interferometric signals in space-borne Gravitational Wave Detectors are measured by digital phasemeters. The phasemeter processes signals generated by multiple interferometers, with its primary function being micro-radian level phase measurements. The Science Interferometer is responsible for inter-spacecraft measurements, including relative ranging, absolute ranging, laser communication, and clock noise transfer. Since the scientific interferometer incorporates multiple functions and various signals are simultaneously coupled into the heterodyne signal, establishing a suitable evaluation environment is a crucial foundation for achieving micro-radian level phase measurement during ground testing and verification. This paper evaluates the phase measurement noise of the science interferometer by simulating the heterodyne signal and establishing a test environment. The experimental results show that when the simulated heterodyne signal contains the main beat-note, upper and lower sideband beat-notes, and PRN modulation simultaneously, the phase measurement noise of the main beat-note, upper and lower sideband beat-notes all reach 2π μrad/Hz^1/2@(0.1 mHz–1 Hz), meeting the requirements of the space gravitational wave detection mission. An experimental verification platform and performance reference benchmark have been established for subsequent research on the impact of specific noise on phase measurement performance and noise suppression methods. Full article

Catalpa bungei C.A.Mey is an economically significant deciduous tree valued for timber production and landscaping applications. An efficient regeneration system is crucial for clonal propagation and serves as a foundation for future molecular breeding in C. bungei. This study established two in vitro regeneration pathways—indirect somatic embryogenesis and shoot organogenesis utilizing mature zygotic embryos as explants. Primary callus was induced from cotyledon, hypocotyl, and plumule explants. A high frequency (45.73%) of yellow-green compact callus was achieved on De-Klerk and Walton (DKW) medium supplemented with 2.0 mg/L 6-BA, 1.0 mg/L zeatin (ZT), and 0.1 mg/L NAA. Subsequent transfer to 1.5× Murashige and Skoog (MS) medium containing 1.5 mg/L 6-BA, 0.2 mg/L ZT, and 0.1 mg/L NAA yielded the highest embryogenic callus induction rate (16.67%). Embryogenic callus demonstrated bipotent potential, generating both adventitious shoots and somatic embryos under specific hormonal conditions. Histological analyses confirmed the typical developmental stages of somatic embryos, from globular to cotyledonary forms, validating the embryogenic origin of regenerated structures. Furthermore, hormone or osmotic additives such as abscisic acid (ABA), Phytagel, and polyethylene glycol 4000 (PEG4000) significantly enhanced somatic embryo induction, with Phytagel at 5.0 g/L achieving the highest rate (76.31%). For shoot organogenesis, the optimal hormonal combination of the 0.6 mg/L 6-BA, 0.4 mg/L KT, and 0.15 mg/L NAA achieved the highest bud induction rate (88.89%) and produced an average of 4.07 adventitious buds per explant. This study presents an efficient regeneration system for C. bungei, providing a practical platform for large-scale propagation and basis for biotechnological applications in woody plants. Full article

Acute lymphoblastic leukemia (ALL) represents the most common type of cancer in the pediatric population. The (1;19)(q23;p13) translocation is a primary chromosomal abnormality present in 3–12% of ALL cases. The current study aims to develop a label-free innovative nanodevice for the ultrasensitive diagnosis of the TCF3-PBX1 chimeric oncogene, featuring simplified operation and rapid analysis using minimal sample volumes, which positions it as a superior alternative for clinical diagnostics and early leukemia identification. The biosensor system was engineered on a nanostructured platform composed of polypyrrole (PPy) and a novel chemically functionalized hybrid nanocomposite of platinum nanospheres and titanium dioxide nanoparticles (TiO₂@Pt). Single-stranded oligonucleotide sequences were chemically immobilized on the nanoengineered transducer to enable biospecific detection. Cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), ultraviolet-visible spectroscopy (UV-Vis), and atomic force microscopy (AFM) were used to characterize each stage of the biotechnological device fabrication process. The analytical properties of the sensing tool were explored using recombinant plasmids containing the TCF3-PBX1 oncogenic sequence and clinical specimens from pediatric patients with B-cell ALL. After exposing the molecular monitoring system to the genetic target, significant variations were observed in the voltammetric oxidation current (∆I = 33.08% ± 0.28 to 124.91% ± 17.08) and in the resistance to charge transfer (ΔR_CT = 19.73% ± 0.96 to 83.51% ± 0.84). Data analysis revealed high reproducibility, with a relative standard deviation of 3.66%, a response range from 3.58 aM to 357.67 fM, a detection limit of 19.31 aM, and a limit of quantification of 64.39 aM. Therefore, a novel nanosensor for multiparametric electrochemical screening of the TCF3-PBX1 chimeric oncogene was described for the first time, potentially improving the quality of life for leukemic patients. Full article

Early and accurate identification of brain tumors is essential for determining effective treatment strategies and improving patient outcomes. Artificial intelligence (AI) and deep learning (DL) techniques have shown promise in automating diagnostic tasks based on magnetic resonance imaging (MRI). This study evaluates the performance of four pre-trained deep convolutional neural network (CNN) architectures for the automatic multi-class classification of brain tumors into four categories: Glioma, Meningioma, Pituitary, and No Tumor. The proposed approach utilizes the publicly accessible Brain Tumor MRI Msoud dataset, consisting of 7023 images, with 5712 provided for training and 1311 for testing. To assess the impact of data availability, subsets containing 25%, 50%, 75%, and 100% of the training data were used. A stratified five-fold cross-validation technique was applied. The CNN architectures evaluated include DeiT3_base_patch16_224, Xception41, Inception_v4, and Swin_Tiny_Patch4_Window7_224, all fine-tuned using transfer learning. The training pipeline incorporated advanced preprocessing and image data augmentation techniques to enhance robustness and mitigate overfitting. Among the models tested, Swin_Tiny_Patch4_Window7_224 achieved the highest classification Accuracy of 99.24% on the test set using 75% of the training data. This model demonstrated superior generalization across all tumor classes and effectively addressed class imbalance issues. Furthermore, we deployed and benchmarked the best-performing DL model on embedded AI platforms (Jetson AGX Xavier and Orin Nano), demonstrating their capability for real-time inference and highlighting their feasibility for edge-based clinical deployment. The results highlight the strong potential of pre-trained deep CNN and transformer-based architectures in medical image analysis. The proposed approach provides a scalable and energy-efficient solution for automated brain tumor diagnosis, facilitating the integration of AI into clinical workflows. Full article

To address the reliability requirements for refrigerated container transport in the cold chain, this study established an experimental platform for phase change cold storage balls. A two-dimensional axisymmetric simplified heat transfer model of the three-dimensional cold storage ball was developed. The reliability of the model was verified through charging experiments. While ensuring a certain level of accuracy (average error less than 10%), the model significantly improved computational efficiency (completing calculations in only 49 s), offering a practical reference value. Based on the established 2D axisymmetric simplified heat transfer model, this study focused on the influence of secondary coolant (ethylene glycol solution) parameters on the charging performance. The results indicate that a smaller diameter of the cold storage ball and a higher flow rate lead to a higher freezing rate of the ball. Under the conditions set in this study, the optimal diameters were determined to be 80 mm and 60 mm, and the optimal inlet flow rate was 3.917 m³/h. This simplified model can provide a reference for the optimal design of phase change cold storage systems in refrigerated containers. Full article

The provision of food in pre-Inca/Inca cultures (1000 BC–≈1532 AD) in environments near Lake Titikaka (approximately 4000 m above sea level) was possible through an agricultural technique called “Waru-Waru”, which consists of filling the space (volume) between rows of land containing plants that are cultivated (a series of earth platforms surrounded by water canals) with water, using water as thermal energy storage to store energy during the day and to regulate the temperature of the soil and crop atmosphere at night. The problem is that these cultures left no evidence in written documents that have been preserved to this day indicating the mathematical models, the physics involved, and the experimental part they performed for the research, development, and innovation of the “Waru-Waru” technique. From a review of the existing literature, there is (1) bibliography that is devoted to descriptive research (about the geometry, dimensions, and shapes of the crop fields (and more based on archaeological remains that have survived to the present day) and (2) studies presenting complex mathematical models with many physical parameters measured only with recently developed instrumentation. The research objectives of this paper are as follows: (1) develop a mathematical model that uses finite differences in fluid mechanics, thermodynamics, and heat transfer to explain the experimental and theory principles of this pre-Inca/Inca technique; (2) the proposed mathematical model must be in accordance with the mathematical calculation tools available in pre-Inca/Inca cultures (yupana and quipu), which are mainly based on arithmetic operations such as addition, subtraction, and multiplication; (3) develop a mathematical model in a sequence of steps aimed at determining the best geometric form for thermal energy storage and plant cultivation and that has a simple design (easy to transmit between farmers); (4) consider the assumptions necessary for the development of the mathematical model from the point of view of research on the geometry of earth platforms and water channels and their implantation in each cultivation area; (5) transmit knowledge of the construction and maintenance of “Waru-Waru” agricultural technology to farmers who have cultivated these fields since pre-Hispanic times. The main conclusion is that, in the mathematical model developed, algebraic mathematical expressions based on addition and multiplication are obtained to predict and explain the evolution of soil and water temperatures in a specific crop field using crop field characterization parameters for which their values are experimentally determined in the crop area where a “Waru-Waru” is to be built. Therefore, the storage of thermal energy in water allows crops to survive nights with low temperatures, and indirectly, it allows the interpretation that the Inca culture possessed knowledge of mathematics (addition, subtraction, multiplication, finite differences, approximation methods, and the like), physics (fluids, thermodynamics, and heat transfer), and experimentation, with priority given to agricultural techniques (and in general, as observed in all archaeological evidence) that are in-depth, exact, practical, lasting, and easy to transmit. Understanding this sustainable energy storage technique can be useful in the current circumstances of global warming and climate change within the same growing areas and/or in similar climatic and environmental scenarios. This technique can help in reducing the use of fossil or traditional fuels and infrastructure (greenhouses) that generate heat, expanding the agricultural frontier. Full article

Cortisol is a key biomarker for stress detection, and its levels can be monitored using point-of-care devices with sensors such as nanoparticles and interdigitated array electrodes (IDEs). This study developed an IDE platform using barium titanate (BaTiO₃) particles synthesized via colloidal precipitation with titanium tetraisopropoxide, barium chloride, and Pluronic^® P123. The calcination temperatures varied between 160 °C and 340 °C, with optimal results observed at 160 °C. Scanning electron microscopy revealed particles with an average size of 26 nm, and Fourier transform infrared spectroscopy confirmed the molecular composition after the removal of P123. X-ray diffraction analysis revealed anatase and brookite phases. Brunauer-Emmett-Teller analysis indicated changes in pore morphology, with samples treated at 160 °C exhibiting a type IV(a) mesoporous structure, a surface area of 163 m²/g, and an average pore diameter of 5.24 nm. Higher temperatures led to transitions to type IV(b) at 260 °C and type V at 340 °C, with reduced pore size. Electrochemical impedance spectroscopy was employed to evaluate the performance of the IDE sensor integrated with BaTiO₃ nanoparticles and albumin across cortisol concentrations ranging from 5.0 to 20 ng/mL. Impedance measurements revealed a significant decrease in impedance (Z′) with increasing cortisol concentrations, indicating increased conductivity. Specifically, Nyquist plots for a saliva sample containing 5 ng/mL cortisol—within the typical physiological range—exhibited a marked increase in charge-transfer resistance (Rct), confirming the sensor’s ability to detect low hormone levels in biological fluids. These findings underscore the potential of BaTiO₃-based IDE platforms at 160 °C for stress biomarker monitoring. Full article