Search Results (44,190)

Background: Multimodal large language models (MLLMs) show promise in medical imaging, but their performance is highly dependent on prompt engineering. This study systematically evaluates how different prompting strategies affect diagnostic accuracy in clinical laboratory image interpretation. Methods: We evaluated five MLLMs (ChatGPT-4o, Gemini 2.0 Flash, Claude 3.5 Sonnet, Grok-2, and Perplexity Pro (Claude 3.5 Sonnet)) using 177 proficiency testing images across three domains: blood smears (n = 78), urinalysis (n = 50), and parasitology (n = 49). Three prompting approaches were compared: (1) complex multi-choice prompts with 20 diagnostic options, (2) zero-shot open-ended prompts, and (3) two-step descriptive-reasoning prompts. Images were sourced from the Taiwan Society of Laboratory Medicine external quality assurance archives with expert consensus diagnoses. Results: Zero-shot prompting significantly outperformed complex multi-choice prompts across all models and domains (p < 0.001). With zero-shot prompts, Gemini achieved 78.5% overall accuracy (urinalysis: 92.0%; parasitology: 75.5%; blood smears: 64.1%), representing a 17% improvement over complex prompts. Two-step descriptive-reasoning prompts further improved blood smear accuracy by 8–12% for top-performing models, but showed minimal benefit in urinalysis and parasitology. The re-query mechanism (“please reconsider”) improved urinalysis accuracy by 7.6% but had a negligible effect on blood smears and parasitology. Conclusions: Prompting strategy critically determines MLLM diagnostic performance. Zero-shot approaches with minimal constraints consistently outperform complex multi-choice formats. The remarkable performance of general-purpose models in structured domains like urinalysis (>90% accuracy) demonstrates the considerable progress of multimodal AI. However, complex morphological tasks like blood smear interpretation require either specialized prompting techniques or domain-specific fine-tuning. These findings provide evidence-based guidance for optimizing AI integration in clinical laboratories. Full article

Metabolic-dysfunction-associated steatotic liver disease (MASLD) is a major chronic liver disorder that progresses through inflammation and fibrosis to cirrhosis, yet no effective pharmacological therapy is available. Extracellular vesicles (EVs), which are key mediators of intercellular communication, have recently been reported to exert preventative and therapeutic effects in disease models. This study evaluated the oral efficacy of EVs derived from the microalga Chlorella vulgaris (CEVs) in an MASLD mouse model. Male C57BL/6J mice were assigned to a control group (normal diet), an MASLD group (choline- and methionine-deficient high-fat diet; CDHF), or CEV group (CDHF + CEVs). Twelve-week CEV administration did not alter the CDHF-induced reduction in circulating lipid levels or produce an increase in hepatic lipid content. However, CEV treatment significantly suppressed CDHF-induced fibrosis with collagen accumulation and reduced the mRNA expression of fibrosis-related genes, including Col1a1, Acta2, Mmp2, and Timp1. CEVs also significantly downregulated the expression of macrophage-derived inflammatory mediators—Ccl2, Ccr2, Il6 and Il1b—and reduced lobular inflammatory foci. These findings suggest that CEVs attenuate hepatic fibrosis by modulating early inflammation associated with steatosis and inhibiting hepatic stellate cell activation. This study supports the potential of CEVs as a novel oral intervention for slowing MASLD progression. Full article

An edge-based and curve-based rule-driven nose detection framework is designed to improve the reliability of face detection. The designed framework combines quadratic curve fitting with a calibrated scoring mechanism that fuses geometric, photometric, and structural information into a unified model. These stages jointly enforce symmetry consistency, reliable tip position, and clear wing boundaries. Candidate face regions are first refined by skin filtering and ellipse validation, from which a mid-lower facial ROI is framed for nasal candidate extraction. We further incorporate eye/mouth hints (EyeMap/MouthMap) to restrict the region of interest (ROI) to the region below the eyes, above the mouth, and between the two eyes. When a mouth is detected, this ROI refinement supersedes the chrominance-red (Cr) channel trimming; otherwise, we fall back to the Cr channel horizontal projection to detect dominant mouth peaks and trim the lower-lip band, thereby suppressing lip interference. A multi-threshold Canny procedure with histogram projection is employed to collect multiple nose rectangles by selecting various vertical and horizontal peaks under three adaptive threshold scales. Within each rectangle, edge contours are quadratically fitted and categorized into U-shape (nasal base), N-shape (nostril rim), and C-shape (nasal wings), enabling rule-based selection of the base, wings, and nostrils. The fused features are then processed by a calibrated geometric–photometric–structural scoring module that uses YCbCr contrasts and red/black penalties to suppress lip and eye confounders. Experiments with diverse faces and lighting conditions show accurate and stable nose localization, with notably reliable wing fitting and nasal base detection, improving the accuracy of face detection. Full article

The quantum approximate optimization algorithm (QAOA) is an efficient method for solving combinatorial optimization problems in quantum computing. These problems involve finding the best solution from a finite set of possibilities. At its core, the QAOA uses an Ansatz circuit composed of alternating unitary operators, the mixing and problem Hamiltonians, that are controlled by a set of parameters. Its goal is to find the optimal parameters so that the final quantum state of the circuit encodes the problem’s solution. While this parameter optimization is often handled by classical optimizers, including constrained optimization by linear approximations (COBYLA) and Nelder–Mead, these methods frequently present local extrema. Therefore, we developed a layerwise grid search (LGS) method as an alternative. Since a full grid search is too time-consuming, the LGS method significantly reduces the search time while still finding a good solution. To demonstrate its effectiveness, we present experimental results for the max-cut problem, comparing the performance of our LGS method against conventional classical optimizers. Full article

Wound healing research has advanced through nanotechnology-based delivery systems that enhance the stability and therapeutic potential of bioactive compounds. Resveratrol, a natural polyphenol with antioxidant and anti-inflammatory properties, shows promise for wound healing but is limited by poor bioavailability. This study investigates the efficacy of nano-liposome-encapsulated resveratrol in enhancing skin wound repair in adult zebrafish (Danio rerio). Using a laser-based ablation method, precise full-thickness skin wounds were induced and monitored over 50 days. Resveratrol-loaded liposomes were prepared and orally administered via gavage to facilitate systemic exposure. Compared to the control and blank liposome groups, resveratrol liposome treatment significantly accelerated wound closure, achieving earlier healing milestones (25%, 50%, and 75%). The zebrafish model provided a regenerative platform for real-time evaluation of nanomedicine-based therapies. This study demonstrates the wound healing effects of resveratrol and liposomal encapsulation, offering a targeted, systemically administered strategy for advanced systemic healing and highlighting zebrafish as a valuable model for preclinical regenerative medicine research. Full article

Dark fermentation is commonly used for producing biohydrogen as a green hydrogen, which can be used as an alternative to fossil fuels. Pilot-scale studies on continuous biohydrogen production from molasses are still limited. In this study, a 60 L pilot-scale up-flow anaerobic sludge bed (UASB) dark fermentation system was operated continuously for biohydrogen production from molasses. The reactor achieved an average hydrogen production rate of 3.64 L H₂/L-d. Attention was paid to evaluating total sugar, rather than COD alone, as a more appropriate process indicator for substrate conversion and hydrogen production performance. In addition, metabolic pathway characteristics and microbial community structure were examined. The results provide useful pilot-scale operational data for the implementation of fermentative biohydrogen production technology. Full article

Background/Objectives: Sarcopenia and age-related muscle weakness pose significant global health challenges, highlighting the need for innovative and sustainable exercise interventions for older adults. This study developed and evaluated an Interactive Smart Mirror Exercise Program and investigated the factors associated with older adults’ usage intention toward the program based on the Decomposed Theory of Planned Behavior (DTPB). Methods: A cross-sectional survey was conducted with 92 community-dwelling older adults in northern Taiwan. Structural equation modeling was applied to test the proposed framework and examine the relationships among the study variables. Results: The results showed a satisfactory model fit (SRMR = 0.071). Attitude, subjective norms, and perceived behavioral control together explained 41.6% of the variance in usage intention. In addition, perceived usefulness, perceived compatibility, interpersonal influence, and self-efficacy were identified as factors significantly associated with usage intention, both directly and indirectly. Conclusions: These findings might support the applicability of the DTPB framework in explaining older adults’ usage intention toward technology-assisted exercise programs and provide insights for the design and implementation of digital exercise interventions in community settings. Full article

Asthma is a chronic respiratory disease characterized by airway hyperresponsiveness, obstruction, and persistent inflammation, arising from complex interactions among genetic, epigenetic, immune, and environmental factors. To elucidate the stage-specific molecular mechanisms underlying asthma progression, we constructed candidate genome-wide genetic and epigenetic networks (GWGENs) of human cells through large-scale biological database mining. Using a system order detection scheme, false-positive interactions were pruned to identify real GWGENs corresponding to three clinical stages of asthma: quiet, exacerbation, and follow-up. Core GWGENs were subsequently extracted from each real network using the principal network projection (PNP) method to highlight dominant regulatory structures and pathogenic pathways. Based on the inferred core networks, key stage-specific biomarkers were identified and further explored as potential drug targets. Drug–target relationships were investigated by integrating gene expression perturbation profiles from the Connectivity Map (cMap), comprising microarray data for 14,207 genes across 1327 compounds. This network-guided analysis enabled the qualitative design of multi-molecule drug combinations tailored to each disease stage. Our results suggest that asthma onset is associated with reduced innate immunity, increased disease susceptibility, and impaired endothelial barrier recovery influenced by microenvironmental factors such as cigarette smoke and lipopolysaccharides, together with genetic and epigenetic alterations. During the exacerbation stage, enhanced differentiation of T cells toward the T helper 2 lineage contributes to airway inflammation and tissue injury. In the follow-up stage, T helper 1–mediated responses are linked to mucus hypersecretion, airway obstruction, and sustained inflammation. Collectively, these findings demonstrate that a systems-level, network-based framework can uncover stage-specific pathogenic mechanisms of asthma and provide hypothesis-generating insights for network-informed drug repurposing strategies. Full article

This study examines how different dimensions of leisure literacy relate to work–leisure boundary management and work performance in digitally connected workplaces, addressing the problem that leisure may function as either a restorative resource or a source of boundary conflict. Drawing on boundary theory, the study adopts an exploratory case-based survey design using data from 75 employees in a Taiwanese fire safety enterprise, combining self-reports, supervisor evaluations, and organizational records, with findings analyzed through correlation, subgroup comparison, and regression-based analyses. The results indicate differentiated pathways: positive leisure attitude is associated with work–leisure balance and higher self-rated performance, whereas excessive leisure involvement is associated with increased boundary conflict. These performance-related patterns were more consistently observed for self-rated than for supervisor-rated performance, so performance implications should be interpreted with appropriate caution. Leisure knowledge shows a regulatory role primarily in reducing conflict rather than directly enhancing balance. The results further suggest that comparative leisure/work importance conditions these relationships: when work and leisure are valued more equally, leisure literacy relates more directly to performance, whereas under value imbalance, boundary management becomes more salient, linking leisure literacy to work outcomes. Family life-cycle differences were also observed, although these are treated as contextual. Overall, the study suggests that leisure literacy may support sustainable work performance by shaping whether leisure functions more as a resource or as a source of friction. By extending boundary theory to the work–leisure interface, the study highlights boundary regulation as a relevant issue for sustainable human resource management in digitally connected environments, particularly under conditions of blurred work–leisure boundaries. Full article

We developed an automated system to address orchard crop damage caused by Formosan macaques, a problem where traditional deterrent methods have proven to be ineffective. The system integrates an Internet Protocol camera with a You Only Look Once version 5 (YOLOv5) object detection model, which was trained on an augmented 6000-image dataset featuring a simulated monkey puppet in an indoor setting to validate its real-time identification capability through simulation. Upon target detection, a high-power laser, controlled via the Message Queuing Telemetry Transport protocol, is actuated to perform dynamic and non-invasive repelling. A web-based Human–Machine Interface (HMI) is provided, allowing users to remotely monitor and adjust strategies. This system offers a low-cost, highly efficient, and scalable solution for smart agriculture, with potential for expansion to other scenarios requiring a high degree of security and defense, such as warehouses and construction sites. Full article

In sandstorm environments, a large number of suspended particles in the air absorb and scatter light, causing strong color bias, low contrast, and blurred details in images. These degradations reduce the reliability of computer vision applications in surveillance systems, intelligent transportation systems, unmanned aerial vehicle monitoring, and outdoor autonomous driving systems. A complete sandstorm image enhancement method was developed in this study by combining sky detection, color correction, contrast enhancement, and adaptive dark channel prior (ADCP) dehazing. The Lab color space was used to correct the color bias. The L channel was enhanced using normalized gamma correction and contrast-limited adaptive histogram equalization to improve brightness and contrast. Then, the sky region is detected to avoid over-processing, preserving the natural appearance of the sky region. Finally, ADCP is applied to non-sky regions for further dehazing. Experiments show that the proposed method provides better subjective and objective performance compared to other algorithms. Full article

This paper studies the reliability and availability of a k-out-of-(M+S) retrial machine repair system with two-way communication, consisting of M primary components and S warm standby components. The system incorporates the retrial behavior of failed components. When the repairman becomes idle, he initiates outgoing calls after a random period either to failed components in the orbit for repair or to components outside the orbit for preventive maintenance. The main contribution of this study is the incorporation of proactive repairman behavior, which more realistically captures operational practices in certain engineering systems. By employing the matrix analytic method together with a recursive approach, the steady-state probabilities of the system are obtained, and several important performance measures are derived. Furthermore, the Runge–Kutta method is used to evaluate the system reliability and the mean time to failure. A sensitivity analysis is conducted to investigate the effects of key system parameters, supported by numerical experiments and graphical illustrations. Finally, a cost–benefit model is formulated, and a genetic algorithm is implemented to determine the optimal values of the decision variables that minimize the cost–benefit ratio. Full article

Wind energy is an important form of clean energy, and its rational utilization represents a crucial solution for mitigating the energy crisis and global warming. In this study, wind energy potential and its long-term changes in the South China Sea (SCS) are evaluated using ERA5 100 m wind data from 1944 to 2023, validated against ASCAT observations. High wind speeds and high wind power density (WPD) are concentrated southwest of Taiwan and southeast of Vietnam. Annual wind availability exceeds 6457 h across most regions, reaching up to 8283 h in optimal locations. WPD and capacity factor peak in winter (up to 2.4 × 10⁸ Wh·m⁻² and >50% capacity factor), with the most stable conditions occurring in the southwestern Taiwan Strait, southeast of the Pearl River Delta, and the Beibu Gulf. Empirical orthogonal function analysis reveals that the first mode of winter WPD accounts for 65.7% of the total variance, with a statistically significant increasing trend since 1990. The interannual variation in wind energy resources in the SCS during winter is controlled by the combined effects of sea surface temperature (SST) anomalies in the tropical Pacific and the Arctic Barents Sea. Specifically, in the years with strong wind anomalies in the SCS, mega-La Niña-type SST patterns in the tropical Pacific trigger anomalous cyclonic circulation in the SCS and the eastern Philippine Sea, while warm anomalies in the Arctic Barents Sea surface drive a wave-like structure of “anticyclone–cyclone–anticyclone” from Siberia to South China. The coupling of the two systems jointly promotes the strengthening of the South China Sea monsoon, leading to increased wind speeds and elevated WPD in the northern SCS. These findings provide a scientific basis for wind farm siting and long-term operational planning in the region. Full article

In this work, hydrocyclones with a diameter of 45 mm and cone lengths of 85 mm and 110 mm were employed to investigate the classification behavior of silicon carbide particles. Numerical simulations were carried out using FLUENT based on computational fluid dynamics (CFD). The internal flow characteristics were modeled using the Volume of Fluid (VOF) approach for multiphase flow, coupled with the Large Eddy Simulation (LES) turbulence model. Furthermore, the Discrete Phase Model (DPM) was applied to track particle trajectories and analyze their dynamic behavior within the hydrocyclone. The experimental results showed that, under identical inlet pressure conditions, the hydrocyclone with a cone length of 110 mm achieved superior separation efficiency. Increasing the cone length leads to a reduction in cone angle, which contributes to improved classification performance. However, practical design constraints limit the extent to which the cone length can be increased. To further explore this effect, an extended cone geometry of 150 mm was investigated through numerical simulation. The CFD results indicate that a longer cone structure enhances air core stability, prolongs particle residence time, and decreases the probability of particle misclassification. These findings suggest that optimizing cone length is an effective strategy for improving hydrocyclone performance. The novelty of this study lies in the integration of experimental validation and numerical simulation to systematically evaluate both practical and extended cone designs, thereby providing deeper insights into the relationship between structural parameters and separation efficiency. Full article

In practical engineering applications, rotor–bearing systems inevitably exhibit inherent or residual faults such as imbalance and shaft-bow, originating from manufacturing tolerances, thermal deformation, or operational loading. Accurate monitoring of these faults and their evolution is fundamental to the effectiveness of modern prognostics and health management (PHM) frameworks. However, if such inherent faults are not identified at an early stage, substantial deviations in fault diagnosis may occur, thereby compromising the accuracy of subsequent prognostic assessments and maintenance strategies. This study presents a hybrid diagnostic methodology that integrates a physics-based model with neural network techniques to enhance rotor fault diagnosis. A Jeffcott rotor subjected to simultaneous disk imbalance and shaft-bow is used to demonstrate the methodology, and the results proves its superior capability for simultaneous fault identification. Nonetheless, discrepancies between model predictions and experimental results are observed, attributed to the presence of inherent faults within the rotor system. To address this issue, algorithms for inherent fault identification and compensation, supported by experimental verification, are developed. Following compensation, the accuracy in simultaneously diagnosing and estimating the parameters of imbalance and shaft-bow is significantly improved. The proposed methodology is designed for seamless integration into real-time monitoring systems of industrial rotating machinery. Full article