Archive Ouverte INSA Lyon
Not a member yet
75807 research outputs found
Sort by
Usagers fréquents des urgences : fondements théoriques et perspectives d’action
No full textInternational audienceLes usagers fréquents des services d’urgence (UFU) représentent, de par l’utilisation excessive qu’ils ont de ces services, un défi majeur pour les systèmes de santé du monde entier. Cette étude examine les comportements des UFU en France, en explorant les tendances longitudinales et en évaluant les facteurs individuels, organisationnels et territoriaux influençant leur utilisation. D’après les données du Système national de données de santé, les UFU représentaient 16 % des patients des structures d’urgences (SU) en 2021, mais 34 % des visites. Ces patients, qui souffrent fréquemment de troubles psychiatriques, respiratoires ou cardiovasculaires, ont des dépenses de santé considérablement plus élevées que les autres usagers des SU. Contrairement aux attentes, aucune association significative n’a été constatée entre un recours fréquent aux SU et un accès limité aux soins primaires. Nos résultats soulignent la nécessité d’une approche multidimensionnelle, combinant les perspectives individuelles et systémiques, pour améliorer la prestation de soins et optimiser l’allocation des ressources. Dans ce cadre, l’Agence régionale de santé du Grand Est a développé une solution de cartographie dynamique afin de mieux comprendre ce phénomène et de permettre des interventions ciblées. Parallèlement, le dispositif d’accompagnement des usagers multiples des urgences (DAUM) mis en place sur le Grand Nancy utilise des méthodes de détection et des stratégies d’accompagnement sur mesure. Bien que des difficultés persistent concernant l’engagement des patients et la qualité des données, les premiers résultats confirment l’utilité de profils d’intervention individualisés pour répondre aux besoins des UFU. Ces résultats ouvrent la voie à l’élaboration de stratégies efficaces de prévention primaire ciblant les facteurs de risque associés à une fréquentation fréquente des urgences
CROQuant: Complex Rank-One Quantization Algorithm, with Application to Butterfly Factorizations
No full textInjunctions, experiences, reappropriations: an exploration of the practices and trajectories of ‘refusal’ of prosthetic limbs
No full textInternational audienc
Structure-Aided Design of a LuxR-Type Quorum Sensing SuFEx-Based Potential Inhibitor: Covalent or Competitive Inhibition?
No full textInternational audienceNew N-benzoyl-l-homoserine lactone derivatives bearing a meta-fluorosulfonyl or a meta-methylsulfonyl group have been designed, synthesized and evaluated as quorum sensing (QS) inhibitors. Docking simulations involving the structure of several targeted LuxR-type receptors suggested that a sulfonyl substituent on the benzene ring can trigger interactions within the binding site, possibly consistent with either covalent SuFEx reaction targeting a tyrosine residue or competitive interaction with additional hydrogen bonding. Biological evaluation of the two meta- methyl or fluorosulfonyl-benzoyl acylhomoserine lactone (AHL) analogs as LuxR-regulated quorumsensing inhibitors showed a significant effect for the fluorosulfonyl derivative with an IC50 value of 15 ± 2 µM, while the methylsulfonyl was found to be a weak inhibitor. The stability of the fluorosulfonyl derivative was confirmed by kinetic studies based on 19F NMR experiments. Investigations dedicated to defining the mechanism of action, either covalent or competitive, were achieved through experiments including inhibition assays without or with pre-incubation in the bacterial medium, and LC/MS analysis with the ExpR protein. The results strongly suggest that the type of inhibition is a competitive one
Reconfigurable Constant Multipliers: Hardware Models, Optimization Algorithm and Applications
No full textThis paper introduces a novel algorithm for building run-time reconfigurable single constant multipliers based on addition/subtraction, fixed bit-shift, and multiplexing. An exhaustive exploration of a wide design space using a mix of constraint programming, depth-first search, and branch-and-prune techniques ensures that the architectures are optimal in terms of hardware cost within their model. In this work, detailed bit-level cost models, both for ASIC and for FPGA, are defined and validated against actual syntheses. Compared to the state of the art, the proposed approach enables much larger constant sets and also significantly improves the performance of the resulting architectures. An application to quantized neural network inference demonstrates a reduction in multiplier area with no degradation in delay or accuracy
RibPull: Implicit Occupancy Fields and Medial Axis Extraction for CT Ribcage Scans
No full textInternational audienceWe present RibPull, a methodology that utilizes implicit occupancy fields to bridge computational geometry and medical imaging. Implicit 3D representations use continuous functions that handle sparse and noisy data more effectively than discrete methods. While voxel grids are standard for medical imaging, they suffer from resolution limitations, topological information loss, and inefficient handling of sparsity. Coordinate functions preserve complex geometrical information and represent a better solution for sparse data representation, while allowing for further morphological operations. Implicit scene representations enable neural networks to encode entire 3D scenes within their weights. The result is a continuous function that can implicitly compesate for sparse signals and infer further information about the 3D scene by passing any combination of 3D coordinates as input to the model. In this work, we use neural occupancy fields that predict whether a 3D point lies inside or outside an object to represent CT-scanned ribcages. We also apply a Laplacian-based contraction to extract the medial axis of the ribcage, thus demonstrating a geometrical operation that benefits greatly from continuous coordinate-based 3D scene representations versus voxel-based representations. We evaluate our methodology on 20 medical scans from the RibSeg dataset, which is itself an extension of the RibFrac dataset. We will release our code upon publication.</div
Nature-based solutions for water management: Pluridisciplinary state-of-the-art and research needs
No full textInternational audienceNature-based Solutions (NbS) offer a way to preserve, manage and restore ecosystems so as to better meet today's societal challenges, by combining benefits for society and the environment, including biodiversity. They are a response to current climate change-related challenges for water management. However, various barriers exist to the implementation of NbS, such as a lack of appropriation of the concept, as well as needs for knowledge and know-how. Focusing on societal challenges linked to water, we highlight the importance of implementing pluridisciplinary and transdisciplinary projects when trying to implement NbS projects. This requires new approaches in research, practice, and governance. This discussion allows identifying levers for a widespread use of NbS for water management
From emerging LEO satellite constellations to the space cloud: Emulation platforms and orchestration methods
No full textInternational audienceIn the rapidly advancing field of satellite communications, mega-constellations of Low Earth Orbit (LEO) satellites are gaining significant attention from the academic and industrial sectors. Managing these expanding constellations has become increasingly complex, and integrating them with classical cellular networks presents new automation challenges. We envision a Space Cloud in which Multiaccess Edge Computing (MEC) services are deployed within cross-liked space networks to address emerging Non-Terrestrial Networks (NTNs) latency demands. Integrating computation services in orbit will be instrumental in unlocking a Space Cloud that reduces the need to route computation requests to the Internet backbone. This study's first contribution is MeteorNet, an open-source constellation and edge computing emulation platform aimed at assessing the expected performance of future Space Clouds. MeteorNet realistically replicates the behavior of edge computing in a synthetic satellite constellation network hosting onboard containerized servers. The second contribution comprises two innovative edge orchestration strategies based on fuzzy logic and reinforcement learning. These strategies leverage historical data on task loads and processing failures to control the activation of on-orbit edge servers, ensuring efficient resource utilization. A Pareto-efficient analysis of multiple Key Performance Indicators (KPIs) using MeteorNet proves the approach's feasibility in space missions with energy constraints and limited computation resources
Large Language Models and Algorithm Execution: Application to an Arithmetic Function
No full textLarge Language Models (LLMs) have recently developed new advanced functionalities. Their effectiveness relies on statistical learning and generalization capabilities. However, they face limitations in internalizing the data they process and struggle, for instance, to autonomously execute algorithms. In this paper, we investigate the possibility of extending these models' capabilities to algorithm execution through specialized supervised training focused on reasoning decomposition. We introduce a training model called LLM-DAL (Large Language Model - Decompositional Algorithmic Learning), through which we demonstrate that LLMs' ability to perform complex algorithmic inferences and generalize can be significantly improved when the training method is properly designed to guide the model in its learning process