Scientific Publications of the University of Toulouse II Le Mirail
Not a member yet
92205 research outputs found
Sort by
La transition agroécologique dans l’enseignement agricole technique, un enjeu d’éducation au politique
No full textInternational audienc
Cytométrie de flux par rétro-injection optique : une approche basée sur les réseaux de neurones pour la classification de particules et de cellules
No full textNational audienceThe detection and characterization of microparticles in fluids is a fundamental task across biomedicine, environmental science, and chemical engineering. While established single-particle analysis methods, such as flow cytometry, are highly effective, their frequent reliance on fluorescent labeling can alternative particle properties and introduce operational complexity. This limitation has motivated the search for compact, sensitive, and label-free alternatives capable of high-throughput analysis.This thesis investigates Optical Feedback Interferometry (OFI) as a promising technology to address this need. In an OFI system, a single laser diode serves as both the light source and the detector, functioning as an interferometer and enabling a uniquely compact and self-aligned sensor configuration. The primary objective of this thesis is to systematically develop and validate a complete OFI-based pipeline for the label-free detection and classification of single microparticles.The research encompasses the entire workflow, beginning with the design of robust optoelectronic and microfluidic platforms to ensure precise particle interrogation. It then addresses the core challenge of interpreting the complex signals generated as particles transit through the laser’s sensing volume. Particular emphasis is placed on developing methods that are reproducible and resilient to the inherent variability of experimental conditions.A key component of this work involved modeling and characterizing the OFI signal in the context of microfluidic flow. The research demonstrates how particle motion, shaped by hydrodynamic focusing, gives rise to specific Doppler-frequency signatures in the sensor output. Building on this understanding, a novel adaptive algorithm was developed to reliably detect and segment the transient signal bursts associated with individual particle passages. This automated segmentation step was critical, enabling the construction of a consistent and validated dataset of particle events that formed the foundation for subsequent classification studies.The final stage of the research focused on leveraging this curated dataset to develop a robust classification framework. Multiple signal representations were explored to extract the most discriminative information related to particle size, including handcrafted statistical features, time–frequency spectrograms, and raw temporal waveforms. These representations were used to train and evaluate a series of machine-learning models. The results confirm that machine learning, when integrated into a tightly controlled experimental pipeline, can classify particles with high accuracy by identifying subtle patterns in OFI signals that are not easily detectable through simple thresholding approaches.In conclusion, this research demonstrates a complete end-to-end framework for label-free single-particle analysis using optical feedback interferometry. By combining a stable experimental platform with advanced signal processing and machine learning, this work establishes a validated methodology for microparticle size classification.La détection et la caractérisation de microparticules dans les fluides constituent une tâche essentielle en biomédecine, en sciences de l'environnement et en génie chimique. Bien que les méthodes établies d’analyse de particules uniques, telles que la cytométrie en flux, offrent de hautes performances, leur dépendance fréquente au marquage fluorescent peut altérer les propriétés natives des particules et introduire une complexité opérationnelle. Cette limitation a stimulé la recherche d’alternatives compactes, sensibles et sans marquage, capables d’assurer une analyse à haut débit.Cette thèse étudie l’interférométrie par rétro-injection optique (OFI) comme technologie prometteuse pour répondre à ce besoin. Dans un système OFI, une unique diode laser sert à la fois de source lumineuse et de détecteur, fonctionnant comme un interféromètre, ce qui permet une configuration de capteur particulièrement compacte et auto-alignée. L’objectif principal de cette thèse est de développer et de valider de manière systématique un pipeline complet basé sur l’OFI pour la détection et la classification sans marquage de microparticules individuelles.La recherche couvre l’ensemble du flux de travail, depuis la conception de plateformes optoélectroniques et microfluidiques robustes garantissant une interrogation précise des particules. Elle aborde ensuite le défi majeur que représente l’interprétation des signaux complexes générés lors du transit d’une particule à travers le volume de détection du laser. Une attention particulière est portée au développement de méthodes reproductibles et robustes face à la variabilité inhérente aux conditions expérimentales.Une partie essentielle du travail a consisté à modéliser et caractériser le signal OFI dans le contexte d’un écoulement microfluidique. La recherche montre comment le mouvement des particules, influencé par la focalisation hydrodynamique, se traduit par des signatures fréquentielles Doppler spécifiques dans le signal du capteur. Sur cette base, un algorithme adaptatif a été développé pour détecter et segmenter de manière fiable les signaux transitoires correspondant aux passages de particules individuelles. Cette segmentation automatisée a constitué une étape critique, permettant la construction d’un jeu de données cohérent et validé d’événements particulaires, qui a servi de fondement à l’ensemble des études de classification ultérieures.La dernière étape de la recherche s’est concentrée sur l’exploitation de ce jeu de données pour développer un cadre de classification robuste. Différentes représentations du signal ont été explorées afin d’extraire les informations les plus discriminantes relatives à la taille des particules, incluant des caractéristiques statistiques extraites manuellement, des spectrogrammes et des formes d’onde temporelles brutes. Ces représentations ont été utilisées pour entraîner et évaluer une série de modèles d’apprentissage automatique. Les résultats confirment que l’intégration de l’apprentissage automatique dans un pipeline expérimental maîtrisé permet de classifier efficacement les particules avec une grande précision, en identifiant des motifs subtils dans les signaux OFI, difficilement détectables par de simples approches basées sur des seuils.En conclusion, cette recherche démontre la faisabilité et l’efficacité d’un cadre complet pour l’analyse sans marquage de particules uniques par interférométrie par rétro-injection optique. En combinant une plateforme expérimentale stable, un traitement avancé du signal et l’apprentissage automatique, ce travail établit une méthodologie validée pour la classification de la taille des microparticules
Modeling pharyngolaryngeal effectiveness using HRCA and features extraction
No full textInternational audienceSwallowing is a physiological mechanism that allows food and liquids to pass from the mouth to the stomach, while ensuring the safety of the airway. When this process malfunctions, it is known as dysphagia. It is estimated that 43.8% of the adult population suffers from dysphagia and aspiration is the leading cause of death in nursing homes. It is therefore important to detect it as early as possible. Currently, two methods of swallowing assessment are considered to be state of the art: fibreoptic swallowing assessment and videofluoroscopic swallowing study. However, they are either invasive or not used systematically. High Resolution Cervical Auscultation (HRCA) is an alternative method that has been studied for swallowing assessment. It allows a non-invasive study based on recording the sounds and movements of the swallowing process using a medical device consisting of a microphone and an accelerometer placed around the subject's neck. We study in the PhLEs-NID project the evaluation of HRCA devices in order to assess swallowing disorders
2965 Voxel-Based Analysis for Predicting Recurrence in Post-Operative Glioblastoma Using Magnetic Resonance Spectroscopy Imaging: Beyond the Cho/NAA Ratio
No full textInternational audienc
Clinical outcome and deep learning imaging characteristics of patients treated by radio-chemotherapy for a “molecular” glioblastoma
No full textInternational audienceBackground Since 2021, glioblastomas have been classified into two subgroups: classic glioblastomas (histGB), defined as IDH wild-type grade 4 astrocytomas with necrosis and vascular proliferation, showing contrast enhancement (CE) on MRI; and molecular glioblastomas (molGB), characterized by specific alterations (7+/10−, EGFR amplification, TERT mutation). Although not always the case, molGB often lack CE and may mimic low-grade gliomas (LGG), hence complicating the diagnosis. Survival outcomes remain debated. This study aimed to evaluate the response of molGB to standard treatment and assess the ability of machine learning and deep learning to differentiate molGB without CE from LGG on MRI. Methods We retrospectively studied 132 glioblastoma patients treated with radiotherapy and temozolomide, comparing the survival outcomes of histGB and molGB. Artificial intelligence (AI) models were trained using features from MRI FLAIR hypersignal segmentation to distinguish molGB without CE from LGG. Results No significant difference in median overall survival (OS) (20.6 vs 18.4 months, P = .2) or progression-free survival (10.1 vs 9.3 months, P = .183) was observed between molGB and histGB. However, molGB without CE demonstrated improved median OS (31.2 vs 18 months, hazard ratios 0.45). Artificial intelligence models distinguished molGB without CE from LGG, achieving a best-performing ROC AUC of 0.85. Conclusions While patients with molGB and histGB have similar overall survival, patients with molGB without CE appear to have better outcomes. Artificial intelligence models effectively differentiate molGB from LGG, supporting their potential diagnostic utility
EARLY PREDICTION OF AT-RISK STUDENTS WITH MINIMAL DATA: A LEARNER PROFILE MODELING APPROACH
No full textInternational audienceWith the rise of courses in learning platforms, especially since the COVID-19 pandemic, universities face a major challenge: high dropout and failure rates. These courses require autonomy and more responsibility from learners, which can lead to difficulties for some. To face this challenge, it is essential to adopt appropriate strategies to identify and support at-risk students from the beginning of the semester. Although several studies have been conducted to identify these students, most of them focus on data collected at the end of the semester, making real-time intervention impossible. Additionally, the studies done during the semester are faced with a lack of relevant data or an insufficient volume of data for reliable predictions. Finally, a few studies explore the prediction of at-risk students with minimal data at the start of the semester. This work aims to identify key indicators (Background, engagement, and pre-test results) to detect at-risk students by modeling early learner profiles of 157 students and then predicting those at risk with minimal data from the beginning of the semester. Our results indicate that high-achieving students generally maintain their level of success, even if their engagement temporarily decreases. Engagement varies among students and evolves differently depending on learner's profiles. The findings showed that most at-risk students belonged to lower-performing profiles who participated less in the first pre-test. In addition, students who participated in all pre-tests showed better final performance, except for some specific profiles. Finally, our analysis revealed that online engagement is not a strong indicator of high academic success and should be combined with other indicators.</div
Distinct neural representational changes following cross-format number tutoring in children with mathematical difficulties
No full textInternational audienceChildren with mathematical difficulties (MD) often struggle to connect abstract numerical symbols with corresponding nonsymbolic quantities, a foundational skill for mathematical development. We evaluated a 4-week personalized cross-format number (CFN) tutoring program designed to strengthen these symbolic-nonsymbolic mappings in children with MD aged 7-10 years. CFN tutoring was associated with significant improvements in numerical and arithmetic fluency. Neural representational similarity (NRS) analysis revealed that deficient cross-format NRS in children with MD was normalized following tutoring, aligning with pre-tutoring levels of typically-developing (TD) peers. This normalization was most pronounced in parietal and parahippocampal regions known to support quantity and spatial representation. We observed a distinctive pattern of neural plasticity across groups-children with MD showed increased cross-format NRS following tutoring, while TD children showed a decrease-suggesting a nonlinear, skill-dependent plasticity. These findings underscore the need for developmentally tailored interventions to support children with MD through targeted, evidence-based strategies
Le wwoofing en tant qu’alternative au tourisme comme espace de co-construction et de transformation des pratiques agricoles et alimentaires. Le cas du Couserans
No full textWwoofing is defined as a non-commercial exchange in which volunteers share the daily life of small farms, mostly engaged in organic farming, in exchange for room and board. This thesis, part of the TOURALIM 2 programme, aims to examine how wwoofing, as an alternative to tourism, can contribute to transforming agricultural and food practices by creating spaces for co-construction between hosts and wwoofers. Our results show that these co-construction dynamics, particularly present in the organisation of the stay and food, but more limited in agricultural work, promote the transmission of knowledge, transform representations and, sometimes, reorient life trajectories. WWOOFing thus appears to be a lever for rebuilding the links between agriculture, food and society.Le wwoofing se définit comme un échange non marchand dans lequel des volontaires partagent le quotidien de fermes paysannes, majoritairement engagées en agriculture biologique, en échange du gîte et du couvert. Ce mémoire, inscrit dans le cadre du programme TOURALIM 2, vise à interroger la manière dont le wwoofing, en tant qu’alternative au tourisme, peut contribuer à transformer les pratiques agricoles et alimentaires en créant des espaces de co-construction entre hôtes et wwoofeurs. Nos résultats montrent que ces dynamiques de co-construction, particulièrement présentes dans l’organisation du séjour et l’alimentation, mais plus limitées dans le travail agricole, favorisent la transmission de savoirs, transforment les représentations et, parfois, réorientent des trajectoires de vie. Le wwoofing apparaît ainsi comme un levier de recomposition des liens entre agriculture, alimentation et société
Scalable and energy efficient communications in the Internet of Things
No full textNational audienceWe have the privilege of being able to quickly setup a video-call and virtually meet old friends living on the other side of the planet. We have the privilege of getting easily and timely updated about what is going on in the world. More recently, we have been given the privilege of being able to remotely check with few clicks if our plants at home are sufficiently moisturized while we are on vacation. In fact, this is the result of the progress in the development of Internet technologies and of its recent extension, the Internet of Things (IoT). The Internet is based on the availability of a resilient networking infrastructure able to let us communicate and access data anywhere at anytime. The IoT further extends the Internet infrastructure by mostly interconnecting very low power devices able to sense physical phenomena and interact with the environment even in remote and inaccessible areas. This last configuration sets up ! some constraints on the design of smart sensing objects and on the way they get plugged into the Internet. Indeed, such devices are usually fed by batteries, and use the stored energy to exchange data through wireless communications. Given that radio communications are the greatest drivers of power consumption, it is of core importance to build energy-saving communication protocols when targeting IoT networks. At the same time, the growing availability of such smart objects paves the way to the development of a vast gamut of monitoring applications, with a consequent massive increase in data exchanges. Hence, IoT communications must be able to insure acceptable performances also on increasing network scales. As matter of fact, designing scalable and energy-saving IoT networks is challenging due to the inherent antagonism of such features. Indeed, an increasing number of devices requires more signaling, and thus a bigger energy consumption. Vice versa, energy can be saved by! limiting the maximum number of handled devices, but this reduces the number of potential users. Interestingly, such a conflict can be solved by optimizing energy efficiency when designing large-scale IoT networks, e.g., by enabling synchronization, scheduling, threshold-based mechanisms, etc.With this methodological approach in mind, the research herein presented tackles protocol design and performance evaluation of energy efficient wireless communications in IoT networks. Such an investigation has been carried out over different network scales. It starts by summarizing the contribution to the development of low power mesh networks, which are based on multi-hop communications. Then, it extends the analysis to Low Power Wide Area Networks, which are featured by single-hop communications over long distances. Afterwards, it details the advent of Direct-to-Satellite IoT communications, which are based on interactions from the ground to Low Earth Orbit satellites. Finally, it envisages the research ahead
A physical noise model for quantum measurements
No full textInternational audienceIn this paper we introduce a novel noise model for quantum measurements motivated by an indirect measurement scheme with faulty preparation. Averaging over random dynamics governing the interaction between the quantum system and a probe, a natural, physical noise model emerges. We compare it to existing noise models (uniform and depolarizing) in the framework of incompatibility robustness. We observe that our model allows for larger compatibility regions for specific classes of measurements