Portail HAL UA (Université d'Angers)
Not a member yet
58742 research outputs found
Sort by
Chronic liver diseases are often diagnosed at an advanced stage despite identification of risk factors: insights from a french national study
No full textInternational audienc
An optimized zebrafish obesogenic test protocol with an artificial intelligence-based analysis software for screening obesogens and anti-obesogens
No full textInternational audienceObesity is defined as a disease in which abnormal excessive body fat accumulation causes adverse effects on health. One proposed contributing factor to the rise in obesity is the exposure to endocrine disruptors acting as obesogens. Semitransparent zebrafish larvae, with their well-developed white adipose tissue (WAT), offer a unique opportunity for studying the effects of toxicant chemicals and pharmaceuticals on adipocyte dynamics and whole-organism adiposity in a vertebrate model. The work presented here is a detailed optimized zebrafish obesogenic test (ZOT) protocol. The method allows to assess the effects of diet composition, drugs and environmental contaminants, acting as obesogens or anti-obesogens, alone or in combination, on WAT levels in zebrafish larvae. Zootechnical parameter guidelines, including larvae rearing conditions, feeding, and selection of larvae to be enrolled are provided. An optimized procedure for in vivo staining of adipocyte lipid droplets with Nile Red before and after exposure to compounds is provided to enhance reproducibility. Using suitable subcutaneous WAT locations, a rationally defined guide for wide-field fluorescence microscopy signal acquisition is proposed. The ZOT analysis software (ZOTAS) was developed to enable automated and efficient image data processing by using custom-trained supervised deep-learning models. The present ZOT protocol distinguishes intrinsic variability of the test method from the biological effect measured. It is the basis of a specific, sensitive and robust quantitative in vivo assay for high-throughput screening of compounds and food content that influence adipocyte hyper/hypotrophy. As such, it provides relevant information for environmental as well as human risk and benefit assessments
Qu'est-ce qu'un paysagiste-concepteur ? Ce que disent huit "landscape architects" de leur travail.
No full textInternational audienc
Physics-Guided Approach with Transfer Learning in Vehicle Lateral Dynamics
No full textInternational audienceSimulating long-horizon state variable trajectories using neural network models is a challenging task due to error accumulation from auto-regressive predictions. Poor generalization exacerbates this issue, particularly under noisy and disturbed data conditions. This paper addresses these challenges by introducing a robust methodology that combines physics-based modeling with data-driven approaches to improve long-horizon simulation accuracy. Specifically, we propose a custom loss function designed to minimize auto-regressive prediction errors across extended time horizons, ensuring alignment between predicted trajectories and actual state measurements. Additionally, we leverage transfer learning by pretraining on simulated data generated from a dynamical model, where parameters are estimated using Bayesian inference, to incorporate prior physical knowledge into the learning process. Our methodology is demonstrated on the lateral dynamics of a vehicle using real-world noisy measurements, achieving competitive performance compared to traditional methods. This work highlights the effectiveness of combining custom loss design and transfer learning to enhance long-term simulation robustnes
Behavioral and Physiological Stress Reactions During Exposure to a Tsunami in Virtual Reality
No full textInternational audienceIntroduction The simultaneous study of physiological stress and behavioural reactions in individuals evacuating during a disaster situation is rarely conducted. Virtual reality (VR) offers an ecological way to study this phenomenon. Objectives We investigated stress and behavioural responses, as well as the influence of others behaviour, in individuals exposed to a tsunamitype natural disaster simulated in VR. Methods Eighty-eight participants (43 women, 45 men, mean age: 26.8, SD = 9.4) took part in the experiment. Three experimental conditions were defined to study emotional contagion: being alone on the beach; being in the presence of calm virtual agents; or being in the presence of “panicked” ones (random assignment). The sense of presence in the virtual environment was assessed using validated scale to evaluate the effectiveness of the VR system. Emotional/Stress response were measured using physiological indicators (heart rate, electrodermal conductance) collected repeatedly over throughout the scenario. Escape behaviour was assessed by measuring movement speed in the virtual environment. Results The mean presence score was 3.32 out of 6 (SD = 0.90). Random effects mixed models for longitudinal data showed an increase in movement speed as soon as the wave appeared, followed by an increase in stress when the alarm was triggered (electrodermal conductance and heart rate), with no significant differences between conditions with or without virtual agents. Conclusion The results suggest a rational, moderate reaction to danger. Combined with the absence of emotional panic contagion, these findings challenge the widespread media portrayal of disaster victims as reacting with irrational panic.</p
Eaux, ville et santé : constats, enjeux et solutions
No full textEditorialInternational audienceQu’entend-on par l’eau dans la ville ? Quelle importance pour notre santé ? Dans un environnement urbain toujours plus dense et imperméable, une urbanisation croissante et dans un contexte de changement climatique, l’eau reste indispensable à la vie, son importance pour la santé et le bien-être est indiscutable. L’eau doit être considérée sous toutes ses formes : souterraines, superficielles, pluviales, potables, usées, etc., ainsi que dans tous ses usages et ses bienfaits. Elle est en effet source [...
