Browse Articles
Discover research articles across all indexed journals
An eye-popping discovery: early vertebrates had four eyes rather than two
Disentangle-and-aggregate feature learning (DAFNet) for motor bearing fault diagnosis
Abstract To address the issues of parameter redundancy and low computational efficiency in traditional convolutional neural networks (CNNs) for motor bearing fault diagnosis, which are caused by increasing network depth, this paper proposes a Disentangle-and-Aggregate Feature Learning Network (DAFNet). This method is designed to overcome the challenges of model deployment on resource-constrained edge devices. Moving beyond the conventional strategy of simply stacking layers, DAFNet innovatively adopts a hierarchical disentanglement and aggregation mechanism. It utilizes a secondary splitting strategy to disentangle shallow, medium, and deep features, followed by terminal feature fusion to achieve an effective representation of fault information. Experimental results based on the CWRU dataset demonstrate that DAFNet achieves a 100% average accuracy in fault diagnosis while significantly reducing both computational overhead and parameter count. Compared with existing mainstream lightweight models, this method exhibits superior generalization performance and inference speed, providing new theoretical support for the efficient application of deep learning in industrial embedded systems.
MFSD-YOLO: A multi-scale feature detection network for pediatric wrist abnormalities in radiographic images
Timely and accurate diagnosis of wrist abnormalities, especially distal radius and ulna fractures, is critical in children and adolescents, yet complicated by skeletal immaturity, overlapping anatomy, and low-contrast fracture lines. To address these challenges, we propose MFSD-YOLO, a multi-scale detection model for pediatric wrist abnormality analysis. The model integrates a Cross-Stage Partial Progressive Multi-Scale Feature Aggregation (CSP_PMSFA) module inspired by GhostNet that applies lightweight multi-scale convolutions on partial channels with partial convolution and residual connections to reduce redundancy and enhance shallow texture and subtle fracture sensitivity. The Feature Pyramid Shared Convolution (FPSConv) module replaces pooling with shared dilated convolutions to expand the receptive field and capture multi-scale context without added cost. The C2 Bi-Level Routing Attention (C2BRA) module, based on C2PSA, uses regional routing and local enhancement to refine focus on relevant areas while balancing accuracy and speed. The Recursive Gradient Dynamic Feature Pyramid Network (RepGDFPN) optimizes top-down and bottom-up multi-scale fusion, reducing semantic loss and improving robustness. Finally, the Sliding Weight Adaptive Loss (SlideLoss) addresses class imbalance, enhancing detection of rare targets. Evaluated on the GRAZPEDWRI-DX dataset, MFSD-YOLO achieves 69.7% mAP@0.5, representing a 5.3% improvement over the baseline YOLOv11, while maintaining 10.8M parameters and 3.2 ms inference speed. These results validate the model’s effectiveness and its potential for real-world deployment in clinical pediatric radiographic analysis.
NASA’s latest telescope is a feat of early-career leadership
Can evolutionary therapy be applied in non-small cell lung cancer?
Abstract Evolutionary therapy (ET) applies principles of evolutionary biology to steer tumour dynamics and forestall or delay treatment resistance, typically guided by data-driven mathematical models. Our aim is to assess whether ET protocols, and specifically Zhang et al.’s protocol proposed for metastatic castrate-resistant prostate cancer, can be theoretically effective for fast-growing metastatic cancers such as stage IV non-small-cell lung cancer (NSCLC). Using longitudinal tumour-burden data from NSCLC patients treated with erlotinib, we systematically evaluate 26 two-population differential-equation models based on classical tumour-growth dynamics, with varying assumptions about density- and frequency-dependent interactions, pharmacokinetics, and treatment-induced death. Previous work by Yin et al. on the same dataset employed an exponential model that omitted density- and frequency-dependent interactions; although it provided a good fit to tumour-burden data, its structure would theoretically lead to poorer outcomes under ET protocols. In contrast, our analysis identifies the minimal model structure required to reproduce the resistance-driven regrowth observed in NSCLC, with the Gompertzian model featuring log-kill dynamics and both density- and frequency-dependent interactions providing the best fit. In this model, Zhang et al.’s protocol prolonged median time-to-progression to 42.3 months compared with 24.8 months under maximum tolerated dose. These results indicate that ET is theoretically a viable treatment strategy for NSCLC. This study offers a practical framework for assessing ET feasibility using clinical data and supports future clinical translation of ET in NSCLC.
Does cultural identity catalyze behavior? A TAM-based study on museum cultural creative products consumption in China
China’s cultural creative products are highly homogenized and can no longer meet the growing aesthetic needs of Chinese consumers. In the museum sector, the sluggish sales of cultural and creative products have become a pain point in its development. Research on consumers’ perceptions and purchase intentions regarding museum cultural creative products has grown significantly in recent years, emerging as a key trend in museum sustainability and communication effectiveness studies. Existing research predominantly adopts a design perspective, evaluating designs based on aesthetic and functional needs, while neglecting the influence of consumers’ cultural identity. This study investigates the influence of cultural identity on Chinese consumers’ willingness to purchase museum cultural products. To achieve this, we developed a Cultural Identity Model (CIM) based on the Technology Acceptance Model (TAM). Based on the analysis of data from 404 respondents, Chinese consumers do not perceive cultural identity (P = 0.476), ethnic identity (P = 0.482), or religious identity (P = 0.242) associated with museum cultural creative products as influential factors in their purchase decisions. However, they focus on the product attributes of cultural and creative designs (materials, colors, P = 0.042) as well as functional attributes (practicality, usability and durability, P < 0.01), all of which effectively enhance consumers’ purchase intention. Grounded in the cultural identity perspective, this study investigates determinants that actively catalyze purchase behavior toward museum-based cultural and creative products. The findings deliver novel design frameworks and theoretical foundations for the cultural and creative products development. This study provides empirical support for optimizing the design of museum cultural and creative products, and suggests focusing on the integration of functional attributes (such as practicality) and carriers of perceived cultural symbols (such as materials and colors).
These mysterious ridges could be the secret to younger skin
Vicarious body maps bridge vision and touch in the human brain
Abstract Our sensory systems work together to generate a cohesive experience of the world around us. Watching others being touched activates brain areas representing our own sense of touch: the visual system recruits touch-related computations to simulate bodily consequences of visual inputs 1 . One long-standing question is how the brain implements this interface between visual and somatosensory representations 2 . Here, to address this question, we developed a model to simultaneously map somatosensory body part tuning and visual field tuning throughout the brain. Applying our model to ongoing co-activations during rest resulted in detailed maps of body-part tuning in the brain’s endogenous somatotopic network. During video watching, somatotopic tuning explains responses throughout the entire dorsolateral visual system, revealing an array of somatotopic body maps that tile the cortical surface. The body-position tuning of these maps aligns with visual tuning, predicting both preferences for visual field locations and visual-category preferences for body parts. These results reveal a mode of brain organization in which aligned visual–somatosensory topographic maps connect visual and bodily reference frames. This cross-modal interface is ideally situated to translate raw sensory impressions into more abstract formats that are useful for action, social cognition and semantic processing 3 .
Quality and reliability of weight management short videos on TikTok and Bilibili: cross-sectional study
How auditory development affects language acquisition: Influences of socioeconomic status and gestational age at birth
Technological advances in recent decades have intensified the need for strong language and literacy skills, such that deficits in these skills can significantly reduce occupational opportunities and social richness. Nonetheless, the causes of language and literacy deficits remain scarcely understood, so treatment consists mostly of drill on the very skills affected individuals struggle to perform. The purpose of this study was to test two related hypotheses. Hypothesis 1 was that delays in the development of the central auditory pathways greatly constrain acquisition of language skills dependent upon auditory development; these skills primarily involve those that are late emerging, namely phonological sensitivity. A corollary is that language skills that begin emergence early in life are less affected by delays in auditory development; this largely encompasses lexicosyntactic knowledge. Hypothesis 2 was that some conditions heretofore recognized as impacting language acquisition (poverty and premature birth, for the purpose of this study) take their toll at least in part by constraining the timely development of the central auditory pathways. To test these hypotheses, 104 children (5–6 years old) spanning continua of socioeconomic status and gestational age at birth were tested on (1) three measures of suprathreshold auditory functions associated with development of the central auditory pathways, (2) two measures of lexicosyntactic knowledge, and (3) two measures of phonological sensitivity. Results largely supported both hypotheses: Strong relationships were found between suprathreshold auditory functions and language measures, especially phonological sensitivity, and both socioeconomic status and gestational age appeared to exert their influence on language acquisition completely or partly through an effect on auditory function. These results should serve to refocus the search for causes of language and literacy deficits from purely environmental shortcomings to biological determinants, with newly inspired directions for interventional approaches.
Daily briefing: Tumours use neurons as hotline to the brain
A novel method of C-VBQ score and CT-HU as risk predictors for cage subsidence after short level ACDF
Predictors of tobacco smoking among youth in an urban slum in Kampala, Uganda: A cross-sectional study
Background Tobacco use remains a significant public health concern worldwide. In Uganda, the youth use tobacco at almost three times the rate of adults, with those residing in slum areas exhibiting even higher prevalence levels. Since 2015, strict laws regulating public tobacco use have been implemented in Uganda, however, these measures have not led to a significant decline in tobacco consumption among the youths in slums. Objective To assess the predictors of tobacco smoking among youth living in the slum areas of Kampala, Uganda. Methods This was a cross-sectional study. It was conducted in Bwaise slum in Kampala, we recruited 422 youths aged 18–30 years. Households were sampled systematically, and quantitative data were analyzed using STATA version 14. Modified Poisson regression with robust standard errors was used, prevalence ratios (PR) were used to measure the associations. Factors were considered significant if p-values were less than 0.05. Results The prevalence of current tobacco smoking was 52.6% while the prevalence of ever tobacco smoking was 71.6%. Most of the participants (87.4%) knew the health effects of tobacco use. Gender (adj.PR = 1.74[95% CI = 1.41–2.14]) and age (adj.PR = 1.38[95%CI = 1.10–1.74]) were the strongest predictors of tobacco smoking: the prevalence of tobacco smoking was 74% higher among males compared to females and 38% higher among those aged 21−30 years compared to their younger counterparts. Education level (adj.PR = 0.84[95%CI = 0.70–0.9]), and income/= (adj.PR = 0.79[05%CI = 0.64–0.97) were also predictive of tobacco smoking. Knowledge was also a predictor with prevalence being 34%, 29%, 42% higher among those who didn’t know that smoking causes serious illness (adj.PR = 1.34[95%CI = 1.09–1.64]), stroke (adj.PR = 1.29[95%CI = 1.06–1.59]) and lung cancer (adj.PR = 1.42[95%CI = 1.11–1.83]) respectively. Conclusion More than half of the youth smoke tobacco despite awareness of its health effects. These findings call for development and implementation of targeted initiatives that address the unique needs and behaviors of males, aged 21–30 years, individuals of education below secondary level while addressing the knowledge gaps about effects of tobacco smoking on human health.
Correction: Associations of heatwaves and their characteristics with ischaemic stroke hospital admissions
Genetics to Improve Outcomes in Schizophrenia (GENios): A within-case molecular genetic study protocol
Introduction Despite significant progress in understanding the genetic basis of schizophrenia, there is a clear gap in our understanding of the genetics of outcomes in schizophrenia, particularly those prioritised by people with the condition. This has impeded progress towards precision psychiatry for schizophrenia and the improvement of outcomes. As genomic cohorts continue to increase in number, size, diversity, and phenotypic detail, sufficient data are now available to enable within-case studies focusing on the genetic basis of outcomes in schizophrenia. Methods and analysis The GENios programme of research will use existing and new data to undertake large-scale genomic studies of schizophrenia outcomes. The project focuses on four key outcomes that were prioritised for research by individuals with lived experience of schizophrenia and align with priorities identified in the literature: antipsychotic treatment response; hospital admissions; occupational functioning; and social functioning. The aims of this project are to understand the genetic contributions to these outcomes in schizophrenia to (i) advance current understanding of the biological mechanisms that drive variability in outcomes, (ii) highlight novel drug targets, and (iii) identify genomic predictors of outcomes that can be leveraged for precision medicine. Ethics and dissemination The GENios project has been granted ethical approval and collaborators are required to have the appropriate ethical permission in place to contribute data to the project. Findings from this study will be interpreted and disseminated with the involvement of lived experience experts in scientific publications and conferences as well as to wider non-scientific communities. Plain English summary Schizophrenia is a severe mental illness that affects how people think, feel, and behave. Whilst current treatments help with some symptoms and effects of the disorder, many people continue to struggle with poor long-term outcomes. Outcomes like repeated hospital stays, difficulties with socialising and relationships, and difficulties with finding and maintaining employment. People with lived experience of schizophrenia have identified some outcomes as research priorities. These research priorities are treatment response, hospital admissions, day-to-day living (e.g. having a job), and having relationships. Many people with schizophrenia are given antipsychotic medications to treat ‘positive’ symptoms (such as unusual (false or disturbing) thoughts and beliefs, and seeing and hearing things that are not there). These medications work for some people but around 25% to 30% of people with the disorder have symptoms that are not helped by standard antipsychotic treatments. These symptoms are classed as treatment-resistant. Many more people find that antipsychotic medications do not help with ‘negative’ symptoms (such as feeling emotionally blank and having trouble with motivation) and difficulty with cognitive abilities (such as concentrating, planning, and memory). This project, called GENios, aims to understand why outcomes differ between people with schizophrenia by looking at their genetics. We understand that social factors play an important role in outcomes but this project is specifically focused on genetics. We will analyse data from people with schizophrenia with an aim to uncover the genetic factors that influence outcomes by using data has already been collected by other research projects. We are carrying out this research in order to identify how genes and knowledge of people’s genetics can help develop healthcare that is more tailored to an individual, so-called precision psychiatry. A plain English glossary of terms is included (S1 Table) to help explain technical language used in the main text.
A randomized controlled trial of artificial intelligence-based analytics for clinical deterioration
Risk assessment in cardiac surgery: Exploring machine learning and laboratory indices as adjunctive tools
Background: Post-operative outcomes of cardiovascular surgery vary greatly among patients for a variety of reasons. While the specific reasons are often multifactorial and complex, certain machine learning methods are promising ways to both estimate mortality after surgery and elucidate important factors linked with mortality. Methods: Using the MIMIC-IV database, we identified 11,261 patients on the cardiovascular surgery unit. We estimated all-cause one-year mortality for each patient after their most recent operation. The dataset included patient demographics (including age, sex, and BMI), as well as the maximum and minimum common laboratory values measured prior to each surgery (including electrolytes, eGFR, and red cell distribution width). Results: Of the models tested, logistic regression outperformed all other approaches with respect to accuracy ( p = 0.0075 with respect to a two-tailed t -test with the next strongest model). The model had an accuracy, sensitivity, and specificity of 85.07%, 82.89%, 85.19% respectively. Furthermore, features weighted heavily by the model are consistent with known predictors of mortality in the literature. Conclusion: Pre-operative laboratory values are effective predictors of all-cause one-year mortality post-cardiovascular surgery used in conjunction with machine learning. Renal function, red cell distribution width, leukocytosis, and erythrocyte indices appear to be important prognostic factors.
Effect of sonication and protease inhibitors on Elisa quantification of selected proteins in bovine udder tissue homogenates
Abstract The protocols for preparing ELISA samples vary in terms of the use of mechanical homogenization (MH) alone or the additional use of protease inhibitors (PI) and/or sonication (S). It is unclear whether the type of protocol can affect the final protein yield; thus, we evaluated how different preparation methods and their combination affect the alpha-casein (CSN1), lactoferrin (LTF), and alkaline phosphatase (ALP) concentrations in udder parenchyma using a sandwich ELISA, and tested whether the udder health state (coagulase-positive staphylococcal infection) affects the protocol efficacy. Samples from 22 cows (11 healthy, 11 infected) were processed using the four protocols. PI and S reduced protein concentrations, with the combination of MH + PI+S giving the lowest values for all three proteins. Unlike CSN1, ALP was the most sensitive to processing. CSN1 levels were higher in infected samples, while lactoferrin and ALP were not influenced by infection. All preparation methods affected the protein levels, irrespective of health state. The highest protein concentrations were obtained using MH alone. Our findings highlight the need for careful optimization and standardization of sample preparation to ensure reliable and reproducible ELISA results.
Effects of time delay on excited quarter- and half-car models with jumping nonlinearities
Nonlinear vehicle dynamics are omnipresent and significantly affect driving performance and safety. Jumping vehicles, in particular, exhibit strong nonlinearities that can lead to severe vibrations and steering instabilities. This study investigates the dynamics of nonlinear jumping vehicle models with and without time delay to clarify their fundamental characteristics. Quarter- and half-car models with jumping nonlinearity are considered, and delayed feedback is introduced as an active suspension control. Numerical simulations under periodic and random excitations reveal several key findings as follows. Both models exhibit sudden, discontinuous transitions from periodic to chaotic dynamics in their bifurcation diagrams. A general relationship between the time period of periodic motions and the forcing period is also identified and further validated using a forced Duffing oscillator with time delay, confirming its generic nature. With the introduction of delay, both in-phase and out-of-phase motions emerge between the front and rear, even under simultaneous excitation. From a practical standpoint, in-phase motion is undesirable, making the realization of out-of-phase behavior within a specific delay range important for improving ride performance. This study identifies the stability regions of nonlinear quarter- and half-car models using bifurcation diagrams, Lyapunov exponents, and frequency response curves. In addition, the inclusion of a time delay is shown to effectively stabilize chaotic motions into periodic responses and to induce both in-phase and out-of-phase oscillations. These findings demonstrate that time delay plays a significant role in enhancing the stability of vehicle models with jumping nonlinearities.