Browse Articles

Chronic disturbance alters seed dispersal traits and frugivores resources in a dry tropical forest

Global burden of drowning and risk factors across 204 countries from 1990 to 2021

Flexible Patched Brain Transformer model for EEG decoding

Abstract Decoding the human brain using non-invasive methods is a significant challenge. This study aims to enhance electroencephalography (EEG) decoding by developing of machine learning methods. Specifically, we propose the novel, attention-based Patched Brain Transformer model to achieve this goal. The model exhibits flexibility regarding the number of EEG channels and recording duration, enabling effective pre-training across diverse datasets. We investigate the effect of data augmentation methods and pre-training on the training process. To gain insights into the training behavior, we incorporate an inspection of the architecture. We compare our model with state-of-the-art models and demonstrate superior performance using only a fraction of the parameters. The results are achieved with supervised pre-training, coupled with time shifts as data augmentation for multi-participant classification on motor imagery datasets.

Gender moderates the mediating effect of psychological capital between physical activity and depressive symptoms among adolescents

Prevalence of pharmaceutical industry conspiracy theories among the polish population

Establishment and characterization of a novel immortalized human aortic valve interstitial cell line

Genetic targeting of myelinated primary afferent neurons using a new NefhCreERT2 knock-in mouse

Abstract Primary afferent neurons that convey somatosensory modalities comprise two large, heterogeneous populations: small-diameter neurons that give rise to slowly conducting unmyelinated axonal C fibers and medium-to-large diameter neurons with fast myelinated A fibers. Despite these two major groupings, tools to differentiate between unmyelinated and myelinated primary afferent fibers by genetic targeting have not been available; in particular, whereas numerous mouse driver lines exist to target different C fiber populations, genetic tools that target myelinated primary afferent populations are scarce. Here we describe a knock-in mouse line expressing tamoxifen-dependent CreERT2 under control of the Nefh gene, which encodes neurofilament heavy chain (NFH or NF200), a protein that is highly enriched in myelinated fibers. This mouse enables highly selective and efficient recombination of Cre-dependent reporters for functional and anatomical interrogation of myelinated fibers while excluding unmyelinated C fibers. In combination with other recombinase-expressing mouse lines, this genetic tool will be valuable for intersectional targeting of subpopulations of myelinated primary afferent fibers.

Development and evaluation of S-carboxymethyl-L-cystine-loaded solid lipid nanoparticles for Parkinson’s disease in murine and zebrafish models

A novel wind speed prediction model based on neural networks, wavelet transformation, mutual information, and coot optimization algorithm

Vorinostat attenuates UVB-induced skin senescence by modulating NF-κB and mTOR signaling pathways

A reference equation for peak oxygen uptake for cycle ergometry in Chinese adult participants

PVA and PVP nanofibers combined with Helichrysum italicum oil preserve skin cell interactions, elasticity and proliferation

Determinants of care-seeking for ARI/Pneumonia-like symptoms among under-2 children in urban slums in and around Dhaka City, Bangladesh

Abstract Childhood pneumonia affects an estimated 18% of under-five children in Bangladesh. Urban slum-dwellers face challenges in healthcare-seeking. This study examined the factors influencing the healthcare-seeking for childhood pneumonia among under-two children in urban slums in Bangladesh. The study examined influence of children’s characteristics (age, sex, number of ARI/pneumonia symptoms, and duration of symptoms), maternal factors (age, education, and working status), and household characteristics (number of household members, wealth quintile, sex of household heads, age of household heads). The outcome variable was receiving care from a qualified medical provider for childhood pneumonia or pneumonia-like symptoms within 14 days before the collection of surveillance data. The research utilized data from the Urban Health and Demographic Surveillance System, which included 155,000 people from five slums in Dhaka and Gazipur City Corporation areas. Overall, 753 out of 4,679 (16%) children under two years of age were included in this study, all of whom had ARI/pneumonia-like symptoms. The mean age of these children was 11.4 months, and 50% were boys. Of them, 350 (46%) sought care from local pharmacies, while 37% sought care from medically trained providers. Logistic regression analyses indicated that children with multiple symptoms (AOR: 2.32, 95% CI: 1.71–3.14) and illness duration over seven days (AOR: 2.61, 95% CI: 1.51–4.51) had higher odds of receiving care from a medically trained provider. Higher maternal education compared to no formal education, having five or more household members compared to four or fewer, household heads aged 40–49 years compared to those under 25 years, a longer duration of living in the slum (more than 10 years compared to less than five years), and belonging to the richest wealth quintile compared to the poorest were protective factors for care-seeking from qualified providers. Further research is required to understand the context for designing appropriate interventions and comprehensive policies for improved child health regarding ARI/pneumonia-like symptoms.

Immunoinformatics method to design universal multi-epitope nanoparticle vaccine for TGEV S protein

Correlation between pierced earrings and the prevalence of metal allergies at Tokushima university hospital: a 15-year retrospective analysis

Research on the desalination kinetics of carbon tableting electrodes for capacitive deionization water purification

UV-Vis spectroscopy coupled with firefly algorithm-enhanced artificial neural networks for the determination of propranolol, rosuvastatin, and valsartan in ternary mixtures

Abstract In the present study, a simple, rapid and cost-effective analytical method was developed for the simultaneous determination of three commonly prescribed cardiovascular drugs: propranolol, rosuvastatin and valsartan. The method employed artificial neural networks (ANN) to model the relation between the UV absorption spectra of the drugs and their concentrations. An experimental design of 25 samples was employed as a calibration set, and a central composite design of 20 samples was used as a validation set. The firefly algorithm (FA) was evaluated as a variable selection procedure to optimize the developed ANN models resulting in simpler models with improved predictive performance as evident by lower relative root mean square error of prediction (RRMSEP) values compared to the full spectrum ANN models. Validation of the developed FA-ANN models demonstrated excellent accuracy, precision and selectivity for the quantification of the target analytes as per international conference on harmonisation (ICH) guidelines. Additionally, the greenness, analytical practicality and sustainability of the developed models were assessed using the analytical greenness (AGREE), blue applicability grade index (BAGI) and the red-green-blue (RGB) tools, confirming their environmentally friendly, practical and sustainable nature. This research shed the light on the potential of ANN coupled with UV fingerprinting for the rapid and simultaneous determination of critical cardiovascular drugs posing a significant impact on pharmaceutical quality control and patient monitoring.

An integration of ensemble deep learning with hybrid optimization approaches for effective underwater object detection and classification model

Stochastic reservoir computers

Ultrafast fMRI reveals serial queuing of information processing during multitasking in the human brain

Abstract The human brain is heralded for its massive parallel processing capacity, yet influential cognitive models suggest that there is a central bottleneck of information processing distinct from perceptual and motor stages that limits our ability to carry out two cognitively demanding tasks at once, resulting in the serial queuing of task information processing. Here we used ultrafast (199 ms TR), high-field (7T) fMRI with multivariate analyses to distinguish brain activity between two arbitrary sensorimotor response selection tasks when the tasks were temporally overlapping. We observed serial processing of task-specific activity in the fronto-parietal multiple-demand (MD) network, while processing in earlier sensory stages unfolded largely in parallel. Moreover, the MD network combined with modality-specific motor areas to define the functional characteristic of the central bottleneck at the stage of response selection. These results provide direct neural evidence for serial queuing of information processing and pinpoint the neural substrates undergirding the central bottleneck.