Browse Articles

IQ and socio-occupational functioning in relation to obsessive-compulsive symptoms severity in a clinical sample of adolescents

Colorimetric dual DNAzyme reaction triggered by loop-mediated isothermal amplification for the visual detection of Shiga toxin-producing Escherichia coli in food matrices

Shiga toxin-producing Escherichia coli (STEC) is causing outbreaks worldwide and a rapid detection method is urgently needed. Loop-mediated isothermal amplification (LAMP) has attracted attention in the development of pathogen detection methods; however, current methods for the detection of LAMP amplicon suffer some drawbacks. In this study, we designed a new LAMP method by incorporating peroxidase-mimicking G-quadruplex DNAzyme for a simple colorimetric detection of the LAMP amplicon. As the new method produces LAMP amplicon containing two DNAzyme molecules per amplification unit, the method was termed colorimetric Dual DNAzyme LAMP (cDDLAMP). cDDLAMP was developed targeting 3 common STEC’s virulence genes (stx1, stx2, and eae) that are associated with serious human illnesses such hemorrhagic colitis and hemolytic-uremic syndrome. Immunomagnetic enrichment was used for specific, ultrasensitive, and fast detection of STEC in food samples (leafy vegetables and milk). The sensitivity of cDDLAMP ranged from 1–100 CFU/mL in pure culture to 100–103 CFU/mL in spiked milk, and 104–109 CFU/25g of lettuce. No cross-reaction with other generic E. coli strains and non-E. coli bacteria was observed. The color signal could be observed by the naked eye or analyzed by either UV–Vis spectra or smartphone platforms. Therefore, the cDDLAMP assay is a cost-effective method for detecting STEC strains without expensive machines or extraction methods.

Analysis and secure communication applications of a 4D chaotic system with transcendental nonlinearities

An origami design for metamaterial robots

Examining the impact of premenstrual dysphoric disorder (PMDD) on life and relationship quality: An online cross-sectional survey study

Little is known regarding the impact of premenstrual dysphoric disorder (PMDD) on specific aspects of life quality within the home, such as spousal/partner relationships. Moreover, the impact of PMDD on the partners of those with the condition has not been investigated. Therefore, the present study examined the ways in which PMDD can affect the perceived life and relationship quality of both those with the condition, and their partners. Across two studies, cross-sectional survey methods were used to compare perceived quality of life and relationship quality between PMDD patients (n = 216) and controls (n = 187), and between PMDD partners (n = 92) and controls (n = 59). In both PMDD patients and their partners, perceived quality of life was lower across most domains compared to controls. Additionally, both PMDD patients and their partners reported lower relationship quality compared to controls, for all domains except love and commitment. Our findings indicate that PMDD has a wide-ranging impact on both the affected individual and their partner. Future clinical research should aim to develop PMDD-specific interventions that support both the person with PMDD and their partner.

Sorption and release of small molecules in PDMS and COC for Organs on chip

Research on memory failure prediction based on ensemble learning

Timely prediction of memory failures is crucial for the stable operation of data centers. However, existing methods often rely on a single classifier, which can lead to inaccurate or unstable predictions. To address this, we propose a new ensemble model for predicting CE-driven memory failures, where failures occur due to a surge of correctable errors (CEs) in memory, causing server downtime. Our model combines several strong-performing classifiers, such as Random Forest, LightGBM, and XGBoost, and assigns different weights to each based on its performance. By optimizing the decision-making process, the model improves prediction accuracy. We validate the model using in-memory data from Alibaba’s data center, and the results show an accuracy of over 84%, outperforming existing single and dual-classifier models, further confirming its excellent predictive performance.

I-MPN: inductive message passing network for efficient human-in-the-loop annotation of mobile eye tracking data

GC-MS, LC-MS, and network pharmacology analysis to investigate the chemical profiles and potential pharmacological activities in flower buds and flowers of Lonicera japonica Thunb

Lonicera japonica Thunb. (L. japonica) is an edible-medicinal herb. While the flower buds of L. japonica are commonly utilized for medicinal purposes, the flowers are often overlooked. However, it has been discovered that the flowers contain higher levels of certain active compounds compared to the flower buds. Despite this finding, there have been no reports on the potential differences in pharmacological efficacy between these compounds. Utilizing results from GC-MS and LC-MS, a total of 335 differential compounds were identified, of which 247 complied with Lipinski’s Rule of Five concerning medicinal properties. Among these, 101 compounds were upregulated in the flower buds, while 146 compounds were upregulated in the flowers. Network pharmacology analysis revealed that the upregulated compounds from the flower buds and flowers targeted 143 and 185 core targets, respectively, with 116 being duplicates. The core target proteins among the duplicate targets were primarily involved in pathways related to cancer, lipid and atherosclerosis, hepatitis B, proteoglycans in cancer, and Alzheimer’s disease. Meanwhile, the hub target proteins upregulated in the flowers enriched distinct pathways associated with human T-cell leukemia virus 1 infection, focal adhesion, the thyroid hormone signaling pathway, and fluid shear stress and atherosclerosis. Molecular docking results indicated that the upregulated compounds exhibited strong binding affinity to the core targets. This study provides insights into the differences in active components between the medicinal (flower buds) and non-medicinal (flowers) raw materials predicting the mechanisms of action of these active components and establishing a basis for the more rational utilization of L. japonica flowers.

Genome-wide identification of SABATH gene family in soybean relate to salt, aluminum, chromium toxicity

Exploring turn demands of an English Premier League team across league and knockout competitions over a full season

Turns are key performance actions in soccer, but can also induce high mechanical loads resulting in tissue damage or injury. This study aimed to quantify the turn demands of an elite English Premier League soccer team. Turning data were obtained from 49 soccer matches (2022–23 season), from a single team that played 35 Premier League, 5 UEFA Europa League, 5 League Cup and 4 FA Cup matches using Sportlight LiDAR technology. Turns were analysed from 29 players who were categorised in playing position groups: goalkeeper (GK), central defenders (CD), full-backs (FB), central-midfielders (CM), wide-midfielders (WM), central-forwards (CF). Turn categories: high (120–180°), medium (60–119°) and low (20–60°) angled, and very high (>7.0ms-1), high (5.5–7.0ms-1), medium (3.0–5.5ms-1), and low (<3.0ms-1) entry speed (ES) was analysed. Primary findings show, on average, per match, CM performed more total turns (~35), than all other playing positions. Additionally, CM performed significantly more low and medium entry speed and high angled turns than other outfield positions. There were no significant differences between turn frequencies and turn characteristics in different competitions (p >0.05). The turning demands of soccer appear to vary significantly between player position. These findings may help inform position-specific return-to-play protocols, physical preparation strategies, drill design and rehabilitation programmes.

Mexican axolotl optimization algorithm with a recalling enhanced recurrent neural network for modular multilevel inverter fed photovoltaic system

Tracking the evolution and persistence of antibiotic resistance in the human gut

The relationship between screen time, screen content for children aged 1-3, and the risk of ADHD in preschools

Objective This study investigates the relationship between screen time, screen content, and the risk of Attention Deficit Hyperactivity Disorder (ADHD) using data from a large sample. Specifically, it examines how different types of screen content (such as educational videos, cartoon videos, and interactive videos) are associated with the risk of ADHD. The aim is to offer a scientific foundation for the rational management of children’s screen time and screen content. Methods We collected data through a questionnaire survey involving a study population of 41,494 children from Longhua District, Shenzhen City, China. The questionnaire recorded the daily screen time and the type of content viewed by the children at ages 1–3 years and assessed their risk of ADHD using the Strengths and Difficulties Questionnaire (SDQ) at ages 4–6 years. Hierarchical logistic regression analysis, controlling for confounding factors, was employed to explore the associations between screen time, screen content, and ADHD risk. Results In the total sample, 6.7% of the participants had screen time exceeding 60 minutes per day, with educational videos predominant type (63.4%). 16.5% of the participants were identified as being at risk for ADHD. Statistically significant positive associations with ADHD were observed across all categories of screen time (P<0.001). Moreover, as screen time increased, the risk of ADHD also rose (OR1~60 mins/d=1.627, 95%CI=1.460~1.813; OR61~120 mins/d=2.838, 95%CI=2.469~3.261; OR>120 mins/d=3.687, 95%CI=2.835~4.796). Significant positive associations with ADHD were observed across all categories of screen time in the educational videos and cartoon videos. For the educational videos group, the odds ratios were as follows: OR1–60 mins/day=1.683 (95% CI=1.481–1.913), OR61–120 mins/day=3.193 (95% CI=2.658–3.835), and OR>120 mins/day=3.070 (95% CI=2.017–4.673). For the cartoon videos group, the odds ratios were: OR1–60 mins/day=1.603 (95% CI=1.290–1.991), OR61–120 mins/day=2.758 (95% CI=2.156–3.529), and OR>120 mins/day=4.097 (95% CI=2.760–6.081). However, no significant associations with ADHD risk were found for any category of screen time in the interactive videos group (OR1~60 mins/d=0.744, 95%CI=0.361~1.534; OR61~120 mins/d=0.680, 95%CI=0.296~1.560; OR>120 mins/d=1.678, 95%CI=0.593~4.748). Conclusion Increased screen time is associated with a higher risk of ADHD, particularly for educational and cartoon videos, while interactive videos show no significant link. To mitigate this risk, parents and educators should implement strategies such as setting time limits, encouraging breaks, and promoting alternative activities. Future research should focus on longitudinal studies and intervention trials to further explore and address this relationship.

Author Correction: Evaluations of dyadic synchrony: observers’ traits influence estimation and enjoyment of synchrony in mirror-game movements

Quantum communication across a 250-kilometre optical-fibre network

Fake leadership influence on organizational destruction in Higher Education Institutions (HEIs)

The role of leadership in Higher Education Institutions (HEIs) is pivotal. While much research emphasizes the beneficial effects of leadership, the negative impacts of destructive leadership styles remain less explored. This paper investigates the concept of fake leadership - a form of destructive leadership characterized by leaders’ intent to engage in harmful behaviors while maintaining a facade of authenticity - and its influence on organizational destruction in HEIs. Drawing on the Toxic Triangle Framework, this study examines how fake leadership undermines intra-organizational trust and job performance, ultimately fostering systemic inefficiencies and organizational decline. The two-step empirical study was conducted with 529 employees from HEIs across Europe (France, Poland, Spain, and the United Kingdom) to verify the hypotheses. Statistical reasoning was based on linear regression analysis with mediators. The results confirmed that fake leadership significantly and positively influences organizational destruction. This effect is mediated by a decline in intra-organizational trust and job performance. This study contributes to the literature by introducing fake leadership as a distinct destructive leadership style in HEIs, providing a tailored framework for understanding its role in enabling organizational destruction, and underscoring the critical role of diminished intra-organizational trust and job performance in deepening these adverse effects. Practical implications emphasize the need for ethical leadership, rigorous selection processes, and proactive measures to safeguard institutional integrity and resilience.

SegMatch: semi-supervised surgical instrument segmentation

Abstract Surgical instrument segmentation is recognised as a key enabler in providing advanced surgical assistance and improving computer-assisted interventions. In this work, we propose SegMatch, a semi-supervised learning method to reduce the need for expensive annotation for laparoscopic and robotic surgical images. SegMatch builds on FixMatch, a widespread semi-supervised classification pipeline combining consistency regularization and pseudo-labelling, and adapts it for the purpose of segmentation. In our proposed SegMatch, the unlabelled images are first weakly augmented and fed to the segmentation model to generate pseudo-labels. In parallel, images are fed to a strong augmentation branch and consistency between the branches is used as an unsupervised loss. To increase the relevance of our strong augmentations, we depart from using only handcrafted augmentations and introduce a trainable adversarial augmentation strategy. Our FixMatch adaptation for segmentation tasks further includes carefully considering the equivariance and invariance properties of the augmentation functions we rely on. For binary segmentation tasks, our algorithm was evaluated on the MICCAI Instrument Segmentation Challenge datasets, Robust-MIS 2019 and EndoVis 2017. For multi-class segmentation tasks, we relied on the recent CholecInstanceSeg dataset. Our results show that SegMatch outperforms fully-supervised approaches by incorporating unlabelled data, and surpasses a range of state-of-the-art semi-supervised models across different labelled to unlabelled data ratios.

Mindfulness-based stress reduction training supplemented with physiological signals from smartwatch improves mindfulness and reduces stress, but not anxiety and depression

Introduction Mindfulness-Based Stress Reduction (MBSR) helps counteract the negative consequences of stress. An essential aspect of mind-body therapies is learning to be mindful of emotional reactions and bodily sensations, a process defined as interoceptive awareness. This awareness can also be enhanced by providing physiological feedback from a smartwatch. However, the impact of using smartwatch-generated physiological signals during mindfulness training has not been studied yet. The study aims at verifying, whether physiological signals from a smartwatch would support the MBSR. Methods We conducted a mixed-design randomized controlled trial to investigate the effects of MBSR training, with and without monitoring physiological signals via a smartwatch, on mental functioning parameters, with measurements taken at baseline and after 8 weeks. Participants were classified into three groups (N = 72): the MBSR group, the MBSR + smartwatch group, and the control group. Between measurement sessions, two groups of participants were engaged in MBSR training, while the third group did not participate in any training. Results Results showed a significant reduction in subjectively perceived stress levels, eating disorder symptoms, and intrusive ruminations in both groups participating in MBSR, compared to the control group. However, a notable difference emerged between the two MBSR groups: in the group with smartwatches, a significant increase in mindfulness was observed. In contrast, in the MBSR group without smartwatches, there was a significant decrease across multiple stress-related components, including: anxiety, cognitive impairment, addictions, sleep disorders symptoms, behaviors indicating lack of entertainment, and poor functioning. Conclusions The results suggest that supplementing MBSR with monitoring interoceptive signals by a smartwatch enhances mindfulness, and maintains the effect of stress and eating disorders symptoms reduction but does not decrease anxiety nor improve general mental functioning. This imposes the need for further research to investigate mechanisms involved when observing interoceptive signals by a smartwatch.

Ambiguity-aware semi-supervised learning for leaf disease classification

Abstract In deep learning, Semi-Supervised Learning is a highly effective technique to enhances neural network training by leveraging both labeled and unlabeled data. This process involves using a trained model to generate pseudo labels to the unlabeled samples, which are then incorporated to further train the original model, resulting in a new model. However, if these pseudo labels contain substantial errors, the resulting model’s accuracy may drop, potentially falling below the performance of the initial model. To tackle the problem, we propose an Ambiguity-Aware Semi-Supervised Learning method for Leaf Disease Classification. Specifically, we present a per-disease ambiguity rejection algorithm that eliminates ambiguous results, thereby enhancing the precision of pseudo labels for the subsequent semi-supervised training step and improving the precision of the final classifier. The proposed method is evaluated on two public leaf disease datasets of coffee and banana across various data scenarios, including supervised and semi-supervised settings, with varying proportions of labeled data. The results indicate that our semi-supervised method reduces the reliance for fully labeled datasets while preserving high accuracy by utilizing the ambiguity rejection algorithm. Additionally, the rejection algorithm significantly boosts precision of final classifier on both coffee and banana datasets, achieving rates of 99.46% and 100.0%, respectively, while using only 50% labeled data. The study also presents a thorough set of experiments and analyses to validate the effectiveness of the proposed method, comparing its performance against state-of-the-art supervised approaches. The results demonstrate that our method, despite using only 50% of the labeled data, achieves competitive performance compared to fully supervised models that use 100% of the labeled data.