Browse Articles

Study on the mechanical properties and microscopic mechanism of xanthan gum improved red clay

Applicability of the daily hydrological models (GR4J, GR5J, and GR6J) in the South Korean basins

A propensity score matched retrospective study of psychological and behavioral therapy impact on muscle function recovery in Guillain Barré Syndrome

Enhanced glioma semantic segmentation using U-net and pre-trained backbone U-net architectures

Evaluation of deep learning models using explainable AI with qualitative and quantitative analysis for rice leaf disease detection

On the complexity matching and multiscale nonlinear perspective of voice restoration via fat injection laryngoplasty in unilateral vocal fold paralysis

Preparation and evaluation of immediate and modified release tablets contained diphenidol hydrochloride

Hybrid adaptive PID control strategy for UAVs using combined neural networks and fuzzy logic

This paper presents a novel hybrid combined neural network and fuzzy logic adaptive proportional, integral, and derivative(NNPID+FPID) control strategy that integrates neural networks and fuzzy logic for optimizing Unmanned Aerial Vehicle(UAV) dynamics by tuning the gains of a PID controller. The proposed approach leverages the strengths of each technique by applying neural networks to fine-tune the y and ψ states, while fuzzy logic enhances the performance of x, z, ϕ, and θ dynamics. A single-layer neural network with 10 hidden neurons is utilized to adjust PID gains for the y and ψ states using proportional, integral, and derivative errors (ep,ei,ed) as inputs. The weights are updated through a gradient descent algorithm minimizing the mean squared error, with a nonlinear sigmoid activation function ensuring adaptability. Concurrently, fuzzy logic employs heuristic rules to dynamically tune PID gains for the remaining states, based on input errors and their derivatives. Membership functions map inputs to gains to ensure real-time adaptability. The hybrid method outperforms standalone neural network(NNPID) and fuzzy logic(FPID) approaches by significantly improving trajectory tracking performance and overall UAV control efficiency. This work demonstrates the effectiveness of combining neural networks and fuzzy logic to address the multi-dimensional control challenges of UAV systems.

Error recognition of english translation text based on neural network and fuzzy decision tree

In response to the low accuracy and recall of current English translation text error recognition methods, this paper proposes a research on English translation text error recognition based on an improved decision tree algorithm. Firstly, use mutual information to calculate the degree of relevance of entries and annotate the part of speech of English translation texts. Then, by using an encoder and decoder to construct a neural network structure, the neural network is applied to the process of feature extraction in machine English translation, and the softmax function is used for normalization. Finally, a fuzzy decision tree is used to segment each information feature, and combined with the Gini index, error features are classified to achieve English translation text error recognition. Through experiments, it has been proven that the accuracy of the translation text error recognition method proposed in this article remains above 94%, the recall rate remains above 86%, the recognition accuracy is high, and the judgment is reliable. The improved decision tree algorithm exhibits consistent performance across varying data volumes, achieving an accuracy exceeding 93% when handling 15,000 to 50,000 sentences, surpassing comparable state-of-the-art algorithms. As such, the enhanced approach for identifying errors in English translation texts effectively enhances translation quality and demonstrates promising potential for practical applications.

A real-world pharmacovigilance study of FDA Adverse Event Reporting System (FAERS) events for Definity

Background Definity significantly enhances the diagnostic accuracy of echocardiography but raises ongoing safety concerns. This study aimed to explore potential adverse events (AEs) associated with Definity in real-world settings by analyzing data from the U.S. Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS). Methods We retrospectively extracted AE reports from the FAERS database between 2004 and the first quarter of 2024. Disproportionality analyses using reporting odds ratio (ROR), proportional reporting ratio (PRR), Bayesian Confidence Propagation Neural Network (BCPNN), and multi-item gamma Poisson shrinker (MGPS) were conducted to identify signals potentially associated with Definity. Univariate and multivariate logistic regression analyses were performed as sensitivity analyses to assess potential risk factors for Definity-related AEs. Results A total of 4460 reports with Definity as the “primary suspected” drug were included. We identified 104 statistically significant signals at PT level, including common AEs such as back pain and muscle spasms, as well as exploratory signals not listed on the drug label, such as eye movement disorder and renal pain. It is noteworthy that 21 signals did not retain significance following Bonferroni correction. These AEs spanned 25 System Organ Classes (SOCs). In reports with available time-to-onset data, most events occurred within 30 days of administration, while some were reported at longer intervals. Moreover, logistic regression analysis indicated that both gender and body weight were independent risk factors associated with Definity-related AE. Conclusion This study provides exploratory insights into potential AEs associated with Definity based on real-world pharmacovigilance data. While most signals are consistent with known safety profiles, several emerging signals warrant further investigation and clinical awareness. These findings may contribute to ongoing risk management and pharmacovigilance efforts.

LLM-optimized wavelet packet transform for synchronous condenser fault prediction

This paper proposes an innovative approach for predicting faults in synchronous condensers in ultra-high voltage direct current (UHVDC) transmission systems. The framework combines Wavelet Packet Transform (WPT) for intelligent feature extraction with an enhanced Gated Recurrent Unit (GRU) network augmented by multi-head attention mechanisms. WPT is employed for efficient decomposition of fault signals into multiple frequency sub-bands, facilitating the extraction of fault features such as energy, entropy, and statistical moments. By applying Large Language Models (LLM) to WPT, an intelligent feature selection mechanism significantly improves both detection accuracy and processing efficiency. The Multi-Head Attention GRU (MHA-GRU) network architecture is designed to capture complex temporal dependencies in fault signals while maintaining computational efficiency. Comprehensive experimental results demonstrate that our framework consistently outperforms state-of-the-art methods across all performance metrics, including classification accuracy, detection time, and false alarm rate. The system exhibits robust stability under varying load conditions with particularly significant improvements in air-gap eccentricity fault detection. The proposed approach provides a reliable solution for early fault prediction in UHVDC synchronous condensers, enabling timely maintenance intervention before minor issues develop into critical failures.

Gamification-based tele-rehabilitation for physical therapy in patients with Parkinson’s disease: A scoping review

Objective A scoping review was conducted to explore what is known about gamification-based tele-rehabilitation (GBT) to enable physical therapy in home settings for people with Parkinson’s disease (PD). Methods The peer-reviewed literature (OVID Medline, OVID EMBASE, CINAHL EBSCO, and Scopus databases) was searched from January 2010 to May 2024, and 24 articles met the inclusion criteria. The methodological quality of the studies was assessed using the Downs and Black evaluation tool, and levels of evidence were assigned using a modified Sackett scale. Results The majority of the 24 studies were of poor to fair methodological quality (83%), and all used a quantitative design with small sample sizes. The focus of the included studies was to enhance whole-body rehabilitation, with most addressing the upper extremities. Thirteen studies customized their games, whereas others utilized strictly commercial systems (n = 11). Eight studies reported no adverse events while the rest did not report on these. Eight studies indicated that participants maintained high levels of motivation and adherence in home settings. Conclusion GBT has the potential to offer a safe, engaging and effective physical therapy to the PD population in home settings, but further research is warranted to help delineate clearer implementation considerations.

Assessing fecal pollution source in a Northern Michigan Lake using qPCR and a community-based monitoring framework

Implementing quantitative polymerase chain reaction (qPCR) within a community-based research framework expands the scope and scale of community-driven monitoring and research efforts. The increasing accessibility of qPCR technology and methodology has allowed the incorporation of community partners in numerous ways, ranging from sample collection to running qPCR tests. Here, we report on a community-driven study at Crystal Lake in Beulah, MI, in which qPCR was demonstrated to be a more valuable water testing technique than culture-based methods. Historically high levels of the enteric bacteria Escherichia coli in the inlet to Crystal Lake, Cold Creek, as measured by culture-based testing methods, spurred an interest in understanding more about fecal pollution and its source. In this study, we monitored 17 sites in Cold Creek and around Crystal Lake throughout the summers of 2020 and 2021 and used qPCR to assess levels of Enterococcus while source-tracking all samples for human, dog, and Canada goose fecal markers (HF183, DG3 and CG0F1-Bac, respectively). Replicate samples were sent for E. coli culture-based testing. Results showed high fecal contamination (E. coli and Enterococcus) and consistent HF183, DG3 and CG0F1-Bac-positive samples at specific sample sites. Varying degrees of relatedness were found between Enterococcus levels grouped by precipitation amount. Due to the nature of the sampling sites, we hypothesize that human fecal contamination is due to stormwater outflows and septic system influences and not direct human contact with the water. A Cohen’s Kappa analysis between the Enterococcus qPCR test results and E. coli culture-based test results indicated a moderately positive relationship. The historical E. coli dataset, now accompanied by the Enterococcus, HF183, DG3 and CG0F1-Bac data, confirms consistent and elevated levels of fecal pollution in Cold Creek and Crystal Lake that is likely related to human sources with stormwater outflows being a contributor to this contamination.

Universal preference for Korean-type grapho-phonemic systematicity: a cross-cultural study of sound-symbol mapping in English, Chinese, and Korean speakers

Recent studies have revealed that writing systems exhibit systematic relationships between letter shapes and their corresponding sounds, termed ‘grapho-phonemic systematicity’. This systematicity manifests differently across writing systems: Semitic languages maximize systematicity through pixel count, Chinese through perimetric complexity, and Korean through Hausdorff distance. This study investigated whether native speakers of these languages would prefer the type of systematicity found in their respective writing systems. An online survey was conducted with 845 participants (271 British, 308 Chinese, and 266 Korean) who were asked to match novel symbols from archaic writing systems with given sound pairs. Contrary to the hypothesis that participants would prefer their native writing system’s systematicity pattern, all groups showed a stronger preference for Korean-type systematicity, where similar sounds correspond to topologically similar symbols. This unexpected finding suggests that modern humans might universally prefer certain types of symbol-sound mapping, possibly influenced by institutionalized education and formal logic training. Interestingly, Korean participants showed the least preference for Korean-type systematicity, potentially due to their meta-knowledge of Hangul’s intentional design. The study reveals a disconnect between how writing systems historically evolved and what modern humans prefer, suggesting that cognitive processes in symbol-sound mapping might have been shaped by modern educational frameworks. These findings contribute to our understanding of universal cognitive principles in visual-auditory mapping and the influence of cultural and educational factors on writing system preferences.

Incidence and progression of diabetic retinopathy and blindness in Indonesian adults with type 2 diabetes

Objectives To report the incidence and progression rate of diabetic retinopathy (DR) and blindness in Indonesian adults with type 2 diabetes. Methods This was a prospective cohort study of 899 adults aged >30 years with confirmed type 2 diabetes. All participants underwent standardized clinical and eye examinations. Two-field retinal photographs were taken. DR was graded by trained grader masked to participants’ clinical details. We categorized DR as follows: mild, moderate, severe non-proliferative (NPDR), and proliferative DR (PDR). Additional category of vision-threatening DR (VTDR) included severe NPDR or more, or moderate NPDR with clinically significant macular edema. Blindness was defined as visual acuity ≤3/60. At least 1-step progression was categorized as DR progressing. Cox-proportional hazard model was used. Results The incidence and progression of DR were 34.6 and 35.1, and incidence of VTDR and blindness were 24.5 and 8.33/1000 person-years, respectively. Longer diabetes duration was associated with increasing the risk of developing DR (Hazard ratio 1.37 [95% confidence interval 1.10–1.72]) and VTDR (2.00 [1.60–2.50]) and progressing DR (1.50 [1.23–1.84]) in 5 years. Obesity was associated with increased risk of developing DR (1.75 [1.14–2.68]) and the presence of gangrene (2.54 [1.49–4.35]) and neuropathy (1.50 [1.07–2.10]) at baseline increased the risk of progressing DR in 5 years. Living in rural area was associated with increased risk of blindness (2.50 [1.03–5.88]). Conclusions We reported the incidence and progression rate of DR, VTDR, and blindness, and documented that longer diabetes duration increased the risk of DR and VTDR in 5 years.

surveydown: An open-source, markdown-based platform for programmable and reproducible surveys

This paper introduces the surveydown survey platform. With surveydown, researchers can create surveys that are programmable and reproducible using markdown and R code, leveraging the Quarto publication system and R Shiny web framework. While most survey platforms rely on graphical interfaces or spreadsheets to define survey content, surveydown uses plain text, enabling version control and collaboration via tools like GitHub. The package renders surveys as interactive Shiny web applications, allowing for complex features like conditional skip logic, dynamic question display, and complex randomization. The package supports a diverse set of question types and formatting options and users can leverage Shiny’s powerful reactive programming model to create a wide variety of interactive features. As an open-source platform, surveydown provides researchers full control over their survey implementation, including the survey application as well as where and how the resulting response data are stored. Workflows are entirely reproducible and integrate seamlessly with existing workflows for data collection and analysis in R.

Global trends in Alzheimer’s disease and other dementias in adults aged 55 and above (1992–2021): An age-period-cohort analysis based on the GBD 2021

Background Alzheimer’s disease and other dementias (ADOD) are growing global health challenges. While existing studies primarily focus on dementia prevention and management in individuals aged 65 and older, evidence suggests that cognitive decline and pathological changes begin earlier (≥55 years). This study focuses on this younger group to enable earlier risk identification and preventive interventions. Methods This study used GBD 2021 data to extract incidence, prevalence, mortality, and DALYs related to ADOD. Trends from 1992 to 2021 were assessed using the Age-Period-Cohort (APC) model. Future burden from 2022 to 2046 was projected with the Nordpred model and validated using the Bayesian Age-Period-Cohort (BAPC) model. Results From 1992 to 2021, ADOD incidence among individuals aged ≥55 increased by 143.88%. The age-standardized prevalence rate (ASPR) rose from 3,870.6 to 3,975.8 per 100,000. Deaths in 2021 were 1.75 times higher than in 1992. The age-standardized DALY rate was consistently higher in females, while males showed an upward trend (net drift, 0.05). APC analysis revealed the steepest incidence increase in the 60–64 age group, with earlier rises in males. Period effects indicated unfavorable incidence trends in high-middle SDI and middle-SDI regions, and similarly adverse mortality trends in high-middle and low-middle SDI regions. Projections suggest a slight increase in ASIR and ASMR by 2046, with females maintaining higher rates than males. Conclusion The global burden of ADOD among individuals aged 55 years and above remains substantial, particularly in East Asia and among females. Given regional heterogeneity, this study recommends developing and implementing region-specific interventions for more effective improvements.

RETRACTED: Prioritizing robots in intelligent manufacturing using q-rung orthopair fuzzy decision-making method and unknown weight information

The rapid evolution of intelligent manufacturing systems necessitates the integration of advanced robotics to meet increasing demands for productivity, precision, and adaptability. Robots play an indispensable role across a spectrum of operations, from assembly to inspection, directly influencing the efficiency and effectiveness of modern manufacturing environments. Addressing the critical need for enhanced decision-making in technological investments, this study evaluates and ranks various types of robots using an integrated decision-making framework. Utilizing the q-rung orthopair fuzzy set (qROFS) to manage uncertain and subjective expert evaluations, this paper combines entropy and similarity measures to determine expert weight coefficients, reflecting the certainty and support degrees of their opinions. Criteria weights are derived using the full consistency method (FUCOM) for subjective weighting and the criteria importance through intercriteria correlation (CRITIC) method for objective weighting. The comprehensive rankings of the robots are then established using the combined compromise for ideal solution (CoCoFISo) method. A practical case study demonstrates the application of the proposed method. Results from the case study indicate that robots for machine tending rank as the most influential, followed by inspection, assembly, welding, and material handling and packaging robots, showcasing their pivotal roles in enhancing manufacturing productivity and safety. This study not only presents a methodological advancement in handling expert uncertainty but also offers actionable insights for integrating robotic technologies in intelligent manufacturing systems, thereby supporting strategic decision-making and operational optimization.

Optimizing pureed diets via texture analysis: A study on the impact of different energy levels and ingredient ratios on nasogastric tube patency

Objective This study aimed to investigate the impact of different energy levels and ingredient ratios on the nasogastric tube patency of pureed diets, optimizing the formulations to meet the nutritional requirements of elderly nasogastric feeding patients while minimizing tube blockage risk. Methods The study followed the guidelines of the “Chinese Resident’s Balanced Diet Pyramid” and formulated five different energy levels of pureed diets (900 kcal, 1200 kcal, 1500 kcal, 1800 kcal, and 2100 kcal) using natural food groups. The diets consisted of seven major food categories: cereals and tubers, vegetables, meats, milk, oil, salt, and fruits. The liquid formulations for the above energy levels were prepared according to the concentration standards for special medical purpose foods (FSMPs). The maximum injection force required for nasogastric feeding was measured via a texture analyzer. The nutritional components of the pureed diets at different energy levels and ingredient ratios were evaluated via West China Hospital Nutrition Software. Spearman correlation analysis, multiple regression analysis, and random forest models were used to explore the relationships between energy levels, nutritional components, ingredients, maximum injection force, and tube patency. Results The study revealed that as the energy density increased, the maximum injection force of the pureed diets significantly increased (p < 0.05), particularly at the 2100 kcal energy level, where the “rice‒carrot‒beef” formula reached the highest value (117.59 ± 0.26 N), whereas the “FSMP” formula at 900 kcal presented the lowest injection force (9.62 ± 0.20 N). There was a significant difference in the impact of different energy levels and formulations on the maximum injection force (p < 0.05). Spearman correlation analysis indicated that carbohydrate (ρ = 0.736) and dietary fiber (ρ = 0.668) contents were significantly positively correlated with the maximum injection force (p < 0.05). Multiple regression analysis further revealed that carbohydrates were the primary factor influencing the injection force, with a regression coefficient of 0.247 (p < 0.05), suggesting that each additional gram of carbohydrate increased the maximum injection force by approximately 0.247 N, whereas the effects of protein, fat, and dietary fiber were not significant (p > 0.05). All nutritional components (energy (ρ = 0.629), carbohydrates (ρ = 0.621), protein (ρ = 0.582), fat (ρ = 0.547), and dietary fiber (ρ = 0.544)) were significantly positively correlated with tube blockage (p < 0.05). Mann‒Whitney U tests revealed that the energy, carbohydrate, protein, fat, and dietary fiber contents in the tube blockage group were significantly greater than those in the nonblockage group (p < 0.05). With respect to food categories, cereals (ρ = 0.742) and meats (ρ = 0.766) were significantly positively correlated with the maximum injection force (p < 0.05). Specifically, rice (ρ = 0.7886) and sweet potato (ρ = 0.506) were significantly positively correlated (p < 0.05), whereas rice flour (ρ = −0.411) and milk (ρ = −0.690) were significantly negatively correlated (P < 0.05). Moreover, cereals (ρ = 0.615) and meats (ρ = 0.628) were significantly positively correlated with the risk of tube blockage at all energy levels (p < 0.05), with rice (ρ = 0.660) and beef (ρ = 0.153) significantly increasing the risk of blockage, whereas rice flour (ρ = −0.350) and milk (ρ = −0.557) were significantly negatively correlated with the risk of blockage (P < 0.05). The random forest model’s feature importance analysis revealed that carbohydrates (33.33%) and dietary fiber (23.01%) were the most important factors for predicting tube blockage, with an AUC value of 0.91, indicating strong predictive ability. Conclusion This study explores the impact of nutritional components and ingredient characteristics on tube patency and blockage risk in nasogastric pureed diets, revealing key optimization pathways for pureed diet formulations. The energy density and ingredient selection of pureed diets significantly affect tube patency. High-energy diets provide higher nutritional density but significantly increase the injection force and blockage risk. Diet formulations should be optimized by reducing high-viscosity and high-hardness ingredients such as rice and beef, using rice flour to replace rice, and milk as the liquid component. For high-energy demands, the carbohydrate and dietary fiber contents should be controlled to reduce the injection force requirements and blockage risk. The study also developed a five-dimensional blockage risk warning model based on energy, protein, fat, carbohydrate, and dietary fiber (AUC = 0.91), classifying low-, medium-, and high-risk levels. Low-risk patients (energy≤1400 kcal/d, carbohydrates≤200 g/d, protein≤70 g/d) are recommended to use homemade formulas, whereas high-risk patients (energy≥1601 kcal/d, carbohydrates≥241 g/d, protein≥86 g/d) should use FSMP for full feeding to balance nutritional supply and tube patency. The findings of this study provide both theoretical and practical guidance for optimizing diets for dysphagia patients, emphasizing that adjusting formulations can effectively balance nutritional supply and tube patency, reduce blockage risk, and prevent malnutrition in homemade pureed feed. This has significant implications for reducing nasogastric complications and ensuring the safety of medical procedures.

SCP-DETR: A efficient small-object-enhanced feature pyramid approach for PCB defect detection

Defects generated during PCB manufacturing, transportation, and storage seriously impact the quality and performance of electronic components. However, detection accuracy is limited due to excessive background interference and the small size of defect targets. To alleviate these issues, this paper proposes an improved PCB defect detection method based on RT-DETR, named SCP-DETR. Firstly, to effectively detect small targets, the S2 feature layer is incorporated into the neck feature fusion. While this improves detection capability, it also introduces considerable computational overhead. To mitigate this, we use SPDConv (Space-to-Depth Convolution) to process the S2 feature layer, reducing the computational complexity. The processed S2 feature layer is then fused with the S3 feature layer and higher-level features. Subsequently, we feed these features into a specially designed CO-Fusion module. By embedding our proposed CSPOKM(CSP Omni-Kernel Module) into the original fusion module, the CO-Fusion module effectively learns feature representations from global to local scales, ultimately enhancing small-target detection performance. Finally, downsampling operations are replaced with PSConv(Pinwheel-shaped Convolution), which better accommodates the Gaussian spatial pixel distributions of subtle small targets. Experimental results demonstrate that the proposed method achieves an mAP@0.5 of 97%, surpassing RT-DETR-r18 by 3%, and an mAP@0.5:0.95 of 53.4%, representing an improvement of 2.2%. Additionally, compared with the recently released YOLOv11m, our method improves mAP@0.5 by 5.6%. These results demonstrate the superior performance of the proposed method in PCB defect detection, which holds significant implications for industrial production. The code is available at https://github.com/Yttong-rr/SCPDETR/tree/master.