HAL-Université de Bretagne Occidentale
Not a member yet
75442 research outputs found
Sort by
Advances in imaging techniques for Sjogren's disease.
No full textInternational audienceImaging of salivary glands (SG), particularly Salivary Glands Ultrasonography (SGUS) is increasingly used in patients with suspected Sjogren's disease (SD). SGUS is the first-line imaging modality. Numerous studies have highlighted this non-invasive, non-irradiating, and low-cost imaging modality. The OMERACT group has established a classification of SG structural damage based on B-mode findings, ranging from grade 0 (normal) to 3 (severe structural damage). SGUS abnormalities (≥ grade 2) have been reported in approximately 63 % of patients with SD. More recently, Hocevar et al. described a Doppler-based classification assessing SG parenchymal vascularization, graded from 0 (normal) to 3 (Doppler signals occupying the entire glandular surface) and could be used as a marker of disease activity and as a biomarker of response to therapy. Moreover, SGUS can be useful for looking for complications such as lymphoma. New ultrasound techniques are currently being developed, including elastography for assessing tissue stiffness, analysis of microvascularization using contrast-enhanced ultrasound with microbubbles, and analysis of minor salivary glands using the ultra-high frequency probe. The combination of several US modalities enhances both sensitivity and specificity of the technique, allowing for the development of a comprehensive multimodal imaging approach. Other imaging techniques can be performed for SD, such as MRI of the parotid glands, allowing analysis of the glandular parenchyma ("salt and pepper" appearance), and certain sequences (DWI-MR) should be performed when lymphoma or other tumors are suspected. 18-FDG PET-CT may be useful to detect systemic manifestations or complications in SD and new PET tracers are currently being developed
Quatorze façons de cibler le lymphocyte B dans le traitement du lupus érythémateux systémique
No full textInternational audienc
Can we trust multispectral drone dataset for eLTER variables monitoring? Sensors sensibility and applications
No full textInternational audienc
Surgical patterns of care of pancreatic cancer. A French population-based study
No full textInternational audienceIntroduction: Surgical resection is the standard recommended treatment in localized pancreatic cancer. The benefit of neoadjuvant chemotherapy is still debated. The aim of this population-based study was to describe the pancreatic cancer surgical management.Material and methods: An observational real-world study from the French Network of Cancer Registries sampled 638 pancreatic adenocarcinomas diagnosed in 2019. Characteristics of patients, tumours and recommended and administered treatments were collected. Operability of the patients and resectability of the tumours were described. A multivariate logistic regression was used to identify factors associated with the probability of having surgical resection.Results: Among the 263 (41 %) patients with M0 pancreatic adenocarcinomas, 202 patients (77 %) were considered operable and 157 (60 %) also had a tumour considered resectable. Upfront resection was recommended for 68 % and resection after neoadjuvant chemotherapy for 32 % of these patients. Among operable patients with resectable tumour, 36 % underwent upfront R0 resection, and 15 % achieved R0 resection following neoadjuvant chemotherapy. Eventually, among M0 pancreatic adenocarcinomas, age over 80 years (OR≥80 years vs < 65 years: 0.16 [0.06-0.39], p < 0.001) and WHO performance status over 0 (OR1-2 vs 0: 0.43 [0.24-0.79], p = 0.013) decreased the odds of having resection. R0 surgical resection was achieved in 61 % of patients selected for upfront surgical recommendation, and 29 % of those selected for a prior neoadjuvant chemotherapy.Conclusion: In a non-selected population, one-third of patients with localized pancreatic cancer had a complete R0 surgical resection. Neoadjuvant chemotherapy did not achieve a resection rate similar to that of patients selected for upfront surgical indication
Étude longitudinale des effets d’un programme d’activités physiques en présentiel ou digital sur la qualité de vie et la motivation de patientes atteintes d'endométriose
No full textInternational audienceContexte : L’endométriose est une pathologie chronique encore méconnue malgré sa prévalence mondiale d’environ 1 femme sur 10 [1]. En dépit des bénéfices engendrés par l’activité physique (AP) dans son accompagnement tertiaire [2], la littérature semble indiquer que les femmes atteintes d’endométriose pratiquent moins que les femmes sans endométriose [3]. Compte tenu des spécificités de l’endométriose, telles que l’isolement social et les contraintes de déplacement, les ressources numériques représentent une option prometteuse pour promouvoir la pratique de l’AP [4]. Cette étude vise à comparer l’efficacité de deux modalités d’accompagnement par l’AP : en présentiel et via une application mobile.Méthodes : 71 femmes (22 en présentiel et 49 en distanciel) atteintes d’endométriose participent à un programme d’accompagnement par de l’AP de 3 mois, réparties en deux groupes : présentiel et distanciel via une application mobile. Trois temps de mesures (début du programme ; 15 jours après le début du programme ; à la fin du programme) sont effectués (i.e., qualité de vie, motivation et intention de poursuite). Des analyses statistiques inférentielles (e.g., modèle linéaire mixte) seront effectuées.Résultats : La récolte des données est en cours. Les résultats et conclusions de cette étude seront présentés lors du colloque. Cette étude doit permettre d’identifier la modalité de pratique la plus bénéfique en termes de motivation à la pratique d’AP et au niveau de l’augmentation de la qualité de vie. Nous supposons que la qualité de vie s’améliorera dans les deux conditions de pratique. Concernant la motivation, cette dernière doit tendre vers une motivation intrinsèque
Fast Adaptation of Multi-Cell NOMA Resource Allocation via Federated Meta-Reinforcement Learning
No full textInternational audienceRadio resource allocation in multi-cellular systems, particularly with non-orthogonal multiple access (NOMA), must be carefully optimized based on real-time user and network conditions, such as channel responses, user population, and inter-cell interference patterns, which naturally fluctuate over time. Fixed machine learning models for radio resource allocation often fail to adapt to these dynamic conditions, leading to suboptimal resource allocation. Moreover, such models struggle to handle inputs and outputs of varying dimensions, limiting their scalability and generalization in time-varying resource allocation problems. To address these challenges, we propose a novel multi-cell, multi-subband NOMA radio resource allocation solution that integrates meta-learning and federated learning (FL) with multi-agent reinforcement learning (MARL). Our solution maximizes energy efficiency (EE) by enabling one-shot adaptation to environmental variations and dynamically managing information dimensionality through the instantiation and removal of agents from a pretrained model. Under this framework, power allocation (PA) and subband allocation (SA) are jointly optimized in a two-stage process: the first stage employs a central reinforcement learning (RL) agent to solve the PA subproblem, while the second stage leverages multi-agent meta-RL combined with FL to address the SA subproblem. Evaluation results demonstrate that our solution effectively adapts to dynamic environments, including variations in channel conditions due to path loss and Doppler effects, as well as fluctuations in the user set. Notably, our approach consistently outperforms the benchmark algorithms, highlighting its robustness and superior adaptability
Research on data storage with a focus on AI on the Edge
No full textInternational audienceDeploying ML/AI algorithms on the edge is necessary for applications (e.g., security and surveillance, industrial IoT, autonomous vehicles, healthcare use cases, ...) requiring low latency, data privacy or reduced costs. However, most edge devices are not equipped with powerful memory systems to process such applications. The objective of this presentation is to show some optimization venues to unlock the memory/storage bottleneck of some ML/AI algorithms mainly from a learning perspective to deply them on low-resource devices.The speaker will first present some past contributions on different topics related to storage systems, then he will focus on some edge AI related storage optimization and conclude with some perspectives
Route-Centric Ant-Inspired Memories Enable Panoramic Route-Following in a Car-like Robot
No full textInternational audienceSolitary foraging ants excel at route following using minimal neural resources, Robots don't. Recent biological studies proposed lateralized, nest-centric memories to explain ants' direct visual homing but did not address how ants follow curved visual routes away from their nest. We present a biologically inspired neuromorphic model for one-shot panoramic route learning and continuous route following, implemented on a compact car-like robot, Antcar. We demonstrate that route-centric lateralized memories, inspired by the insect mushroom body, enable Antcar to achieve bi-directional route-following, with motivation-driven recognition of route extremities and familiarity-based velocity control. With rigorous Lyapunov-based stability analysis and an empirical memory scalability evaluation, the model was tested over 1.6 km across 113 challenging real-world trials. The system achieves less than 25 cm median lateral error using minimal resources (800-pixel input, 300 MB RAM, 500 mW power, and 18.75 kB memory per 50 m route), offering insights into insect cognition and advancing autonomous robotics under strict resource constraints
Mesurer le désaccord entre vocabulaires d'experts et son impact sur l'analyse de données
No full textInternational audienceA vocabulary materialized by linguistic variables is a way to integrate subjective and expert knowledge about data to analyze. When multiple experts are involved in the analysis process, it is crucial to point out disagreements between their respective vocabularies and the potential impact of this disagreement on the output of a data analysis task applied to the processed data. This paper presents disagreement measures that inform experts about discrepancies regarding the topological structure of their partitions, the terminology associated with each modality, and the impact of the disagreement on the discretization of a given dataset. Experiments show that the proposed multi-level disagreement measure exhibits complementary and meaningful discrepancies that may help experts converge toward a consensual vocabulary.Un vocabulaire représenté par des variables linguistiques permet d'exploiter des connaissances subjectives et expertes pour analyser des données. Lorsque plusieurs experts sont impliqués, il est crucial de signaler les désaccords potentiels entre leurs vocabulaires respectifs. Cet article propose des mesures pour quantifier ces divergences entre vocabulaires, à la fois concernant la structure de leurs partitions, les étiquettes linguistiques des modalités ainsi que l'impact du désaccord sur le traitement des données considérées. Les expériences montrent que ces indicateurs identifient des divergences complémentaires et significatives entre les vocabulaires des experts
Création et résolution de problèmes en SEGPA
No full textInternational audienceAu sein du groupe IREM, il s'agit de mettre en œuvre une séquence de création-résolution de problèmes avec des élèves de SEGPA. Les séances de classe sont analysées. L'idée est ensuite de produire un document utilisable par d'autres professeurs