Towards Context-Rich Automated Biodiversity Assessments: Deriving AI-Powered Insights from Camera Trap Data.

Camera traps offer enormous new opportunities in ecological studies, but current automated image analysis methods often lack the contextual richness needed to support impactful conservation outcomes. Integrating vision-language models into these workflows could address this gap by providing enhanced contextual understanding and enabling advanced queries across temporal and spatial dimensions. Here, we present an integrated approach that combines deep learning-based vision and language models to improve ecological reporting using data from camera traps.

Control of Overfitting with Physics.

While there are many works on the applications of machine learning, not so many of them are trying to understand the theoretical justifications to explain their efficiency. In this work, overfitting control (or generalization property) in machine learning is explained using analogies from physics and biology. For stochastic gradient Langevin dynamics, we show that the Eyring formula of kinetic theory allows to control overfitting in the algorithmic stability approach-when wide minima of the risk function with low free energy correspond to low overfitting.

Effectiveness and tolerability of lactic acid vaginal gel compared to oral metronidazole in the treatment of acute symptomatic bacterial vaginosis: a multicenter, randomized-controlled, head-to-head pilot study.

Background: Bacterial vaginosis (BV) is a prevalent vaginal condition among reproductive-age women, characterized by off-white, thin vaginal discharge with a fishy odor. It increases susceptibility to sexually transmitted diseases (STDs) and pelvic inflammatory disease (PID). BV involves a shift in vaginal microbiota, with reduced lactobacilli and increased anaerobic bacteria.

A modeling study to define guidelines for antigen screening in schools and workplaces to mitigate COVID-19 outbreaks.

Background: In-person interaction offers invaluable benefits to people. To guarantee safe in-person activities during a COVID-19 outbreak, effective identification of infectious individuals is essential. In this study, we aim to analyze the impact of screening with antigen tests in schools and workplaces on identifying COVID-19 infections.

A Dynamical Analysis of the Alignment Mechanism Between Two Interacting Cells.

In this work we analytically investigate the alignment mechanism of self-propelled ellipse-shaped cells in two spatial dimensions interacting via overlap avoidance. By considering a two-cell system and imposing certain symmetries, we obtain an analytically tractable dynamical system, which we mathematically analyse in detail. We find that for elongated cells there is a half-stable steady state corresponding to perfect alignment between the cells.

Prevalence and determinants of metabolic syndrome in 2338 childhood cancer survivors: A Dutch Childhood Cancer Survivor LATER 2 study.

Background: Because the occurrence of metabolic syndrome (MetS) might contribute to childhood cancer survivor's excess risk of cardiovascular disease, the authors assessed the prevalence and determinants of MetS in the Dutch Childhood Cancer Survivor Study (DCCSS-LATER2) cohort.

Methods: In total, 2338 adult childhood cancer survivors (CCS) were cross-sectionally assessed for the prevalence of MetS, using the Lifelines cohort (N = 132,226 adults without a history of cancer) as references. The prevalence of MetS was clinically assessed using existing classifications, as well as an alternative method using dual-energy x-ray absorptiometry fat% instead of waist circumference to define abdominal adiposity.

Direct specification of lymphatic endothelium from mesenchymal progenitors.

During embryogenesis, endothelial cells (ECs) are generally described to arise from a common pool of progenitors termed angioblasts, which diversify through iterative steps of differentiation to form functionally distinct subtypes of ECs. A key example is the formation of lymphatic ECs (LECs), which are thought to arise largely through transdifferentiation from venous endothelium. Opposing this model, here we show that the initial expansion of mammalian LECs is primarily driven by the in situ differentiation of mesenchymal progenitors and does not require transition through an intermediate venous state.

Sensitive liquid biopsy monitoring correlates with outcome in the prospective international GPOH-DCOG high-risk neuroblastoma RT-qPCR validation study.

Background: Liquid biopsies offer less burdensome sensitive disease monitoring. Bone marrow (BM) metastases, common in various cancers including neuroblastoma, is associated with poor outcomes. In pediatric high-risk neuroblastoma most patients initially respond to treatment, but in the majority the disease recurs with only 40% long-term survivors, stressing the need for more sensitive detection of disseminated disease during therapy.

The impact of cognitive bias about infectious diseases on social well-being.

Introduction: We investigate the relationship between bias, that is, cognitive distortions about the severity of infectious disease and social well-being.

Materials And Methods: First, we establish empirically the existence of bias and analyze some of its causes; specifically, during the COVID-19 pandemic. Second, we derive an integrated economic-epidemiological differential equation model from an agent-based model that combines myopic rational choice with infectious disease dynamics.

Scaling limit of domino tilings on a pentagonal domain.

We consider the six-vertex model at its free-fermion point with domain wall boundary conditions, which is equivalent to random domino tilings of the Aztec diamond. We compute the scaling limit of a particular nonlocal correlation function, essentially equivalent to the partition function for the domino tilings of a pentagon-shaped domain, obtained by cutting away a triangular region from a corner of the initial Aztec diamond. We observe a third-order phase transition when the geometric parameters of the obtained pentagonal domain are tuned to have the fifth side exactly tangent to the arctic ellipse of the corresponding initial model.

Impact of a Clinical Decision Support System on the Efficiency and Effectiveness of Performing Medication Reviews in Community Pharmacies: A Randomized Controlled Trial.

Background: Clinical decision support systems (CDSSs) already support community pharmacists in conducting medication reviews (MRs) by identifying important information on interactions and suggesting clinical solutions. However, their impact in terms of quality and time savings is widely unexplored. The aim of our study was to investigate whether MRs are performed faster and better with or without using a CDSS.

Rethinking the physiological cross-sectional area of skeletal muscle reveals the mechanical advantage of pennation.

The shape of skeletal muscle varies remarkably-with important implications for locomotor performance. In many muscles, the fibres are arranged at an angle relative to the tendons' line of action, termed the pennation angle. These pennate muscles allow more sarcomeres to be packed side by side, enabling the muscle to generate higher maximum forces for a given muscle size.

The impact of natural climate variability on the global distribution of Aedes aegypti: a mathematical modelling study.

Background: Aedes aegypti spread pathogens affecting humans, including dengue, Zika, and yellow fever viruses. Anthropogenic climate change is altering the spatial distribution of Ae aegypti and therefore the locations at risk of vector-borne disease. In addition to climate change, natural climate variability, resulting from internal atmospheric processes and interactions between climate system components (eg, atmosphere-land and atmosphere-ocean interactions), determines climate outcomes.

Optimal bailout strategies resulting from the drift controlled supercooled Stefan problem.

We consider the problem faced by a central bank which bails out distressed financial institutions that pose systemic risk to the banking sector. In a structural default model with mutual obligations, the central agent seeks to inject a minimum amount of cash in order to limit defaults to a given proportion of entities. We prove that the value of the central agent's control problem converges as the number of defaultable institutions goes to infinity, and that it satisfies a drift controlled version of the supercooled Stefan problem.

On a novel gradient flow structure for the aggregation equation.

The aggregation equation arises naturally in kinetic theory in the study of granular media, and its interpretation as a 2-Wasserstein gradient flow for the nonlocal interaction energy is well-known. Starting from the spatially homogeneous inelastic Boltzmann equation, a formal Taylor expansion reveals a link between this equation and the aggregation equation with an appropriately chosen interaction potential. Inspired by this formal link and the fact that the associated aggregation equation also dissipates the kinetic energy, we present a novel way of interpreting the aggregation equation as a gradient flow, in the sense of curves of maximal slope, of the kinetic energy, rather than the usual interaction energy, with respect to an appropriately constructed transportation metric on the space of probability measures.

Collective effect of self-learning and social learning on language dynamics: a naming game approach in social networks.

Linguistic rules form the cornerstone of human communication, enabling people to understand and interact with one another effectively. However, there are always irregular exceptions to regular rules, with one of the most notable being the past tense of verbs in English. In this work, a naming game approach is developed to investigate the collective effect of social behaviours on language dynamics, which encompasses social learning, self-learning with preference and forgetting due to memory constraints.

Sort & Slice: a simple and superior alternative to hash-based folding for extended-connectivity fingerprints.

Extended-connectivity fingerprints (ECFPs) are a ubiquitous tool in current cheminformatics and molecular machine learning, and one of the most prevalent molecular feature extraction techniques used for chemical prediction. Atom features learned by graph neural networks can be aggregated to compound-level representations using a large spectrum of graph pooling methods. In contrast, sets of detected ECFP substructures are by default transformed into bit vectors using only a simple hash-based folding procedure.

Causal effect of chemotherapy received dose intensity on survival outcome: a retrospective study in osteosarcoma.

Background: This study aims to analyse the effects of reducing Received Dose Intensity (RDI) in chemotherapy treatment for osteosarcoma patients on their survival by using a novel approach. Previous research has highlighted discrepancies between planned and actual RDI, even among patients randomized to the same treatment regimen. To mitigate toxic side effects, treatment adjustments, such as dose reduction or delayed courses, are necessary.

HeMiTo-dynamics: a characterisation of mammalian prion toxicity using non-dimensionalisation, linear stability and perturbation analyses.

Prion-like proteins play crucial parts in biological processes in organisms ranging from yeast to humans. For instance, many neurodegenerative diseases are believed to be caused by the production of prion-like proteins in neural tissue. As such, understanding the dynamics of prion-like protein production is a vital step toward treating neurodegenerative disease.

Multi-Grid Reaction-Diffusion Master Equation: Applications to Morphogen Gradient Modelling.

The multi-grid reaction-diffusion master equation (mgRDME) provides a generalization of stochastic compartment-based reaction-diffusion modelling described by the standard reaction-diffusion master equation (RDME). By enabling different resolutions on lattices for biochemical species with different diffusion constants, the mgRDME approach improves both accuracy and efficiency of compartment-based reaction-diffusion simulations. The mgRDME framework is examined through its application to morphogen gradient formation in stochastic reaction-diffusion scenarios, using both an analytically tractable first-order reaction network and a model with a second-order reaction.

Uncovering the mechanical secrets of the squirting cucumber.

Rapid movement is rare in the plant kingdom, but a prerequisite for ballistic seed dispersal. A particularly dramatic example of rapid motion in plants is the squirting cucumber () which launches its seeds explosively via a high-pressure jet. Despite intriguing scientists for centuries, the exact mechanism of seed dispersal and its effect on subsequent generations remain poorly understood.