Дисертації з теми "Imagerie intelligente"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся з топ-50 дисертацій для дослідження на тему "Imagerie intelligente".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Переглядайте дисертації для різних дисциплін та оформлюйте правильно вашу бібліографію.
Pincet, Lancelot. "Dynamic excitation systems for quantitative and super-resolved fluorescence microscopy." Electronic Thesis or Diss., université Paris-Saclay, 2024. http://www.theses.fr/2024UPASP033.
Single Molecule Localization Microscopy (SMLM) is a super-resolution optical technique enabling the observation of biological samples labeled with fluorescent dyes at resolutions well below the diffraction limit. The quality of this imaging heavily relies on the ability to observe molecules individually, requiring precise control of fluorescent dye photophysics for them to emit with a high sparsity in both space and time. Until now, dynamic excitation methods aimed to produce uniform illumination over large fields (200 um x 200 um). However, these types of illumination encounter difficulties in imaging dense biological samples, such as neurons, where the diversity in dye density prevented the generation of a uniform single molecule regime across the entire observed area. To address this issue, I propose a new approach that dynamically adjusts illumination based on sample density. This method combines a novel tri-dynamic optical excitation system with a feedback loop based on density analysis, benefiting from an in-depth study of fluorescent dye photophysics. The intelligent imaging system, where the excitation pattern varies over time, integrates a 2D scanning system, a variable zoom system, and a laser. This allows for the generation of a variety of dynamically changing illumination patterns to adapt to the observed sample and the density of locally detected localizations. This new approach has been validated on various biological samples. Additionally, the dynamic excitation system has also been explored for live samples imaging techniques, such as MSIM or FRAP
Filali, Wassim. "Détection temps réel de postures humaines par fusion d'images 3D." Toulouse 3, 2014. http://thesesups.ups-tlse.fr/3088/.
This thesis is based on a computer vision research project. It is a project that allows smart cameras to understand the posture of a person. It allows to know if the person is alright or if it is in a critical situation or in danger. The cameras should not be connected to a computer but embed all the intelligence in the camera itself. This work is based on the recent technologies like the Kinect sensor of the game console. This sensor is a depth sensor, which means that the camera can estimate the distance to every point in the scene. Our contribution consists on combining multiple of these cameras to have a better posture reconstruction of the person. We have created a dataset of images to teach the program how to recognize postures. We have adjusted the right parameters and compared our program to the one of the Kinect
Burbano, Andres. "Système de caméras intelligentes pour l’étude en temps-réel de personnes en mouvement." Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLS139/document.
We propose a detection and trackingsystem of people moving in large spacessystem. Our solution is based on a network ofsmart cameras capable of retrievingspatiotemporal information from the observedpeople. These smart cameras are composed bya 3d sensor, an onboard system and acommunication and power supply system. Weexposed the efficacy of the overhead positionto decreasing the occlusion and the scale'svariation.Finally, we carried out a study on the use ofspace, and a global trajectories analysis ofrecovered information by our and otherssystems, able to track people in large andcomplex spaces
Panaïotis, Thelma. "Distribution du plancton à diverses échelles : apport de l'intelligence artificielle pour l'écologie planctonique." Electronic Thesis or Diss., Sorbonne université, 2023. http://www.theses.fr/2023SORUS155.
As the basis of oceanic food webs and a key component of the biological carbon pump, planktonic organisms play major roles in the oceans. However, their small-scale distribution − governed by biotic interactions between organisms and interactions with the physico-chemical properties of the water masses in their immediate environment − are poorly described in situ due to the lack of suitable observation tools. New instruments performing high resolution imaging in situ in combination with machine learning algorithms to process the large amount of collected data now allows us to address these scales. The first part of this work focuses on the methodological development of two automated pipelines based on artificial intelligence. These pipelines allowed to efficiently detect planktonic organisms within raw images, and classify them into taxonomical or morphological categories. Then, in a second part, numerical ecology tools have been applied to study plankton distribution at different scales, using three different in situ imaging datasets. First, we investigated the link between plankton community and environmental conditions at the global scale. Then, we resolved plankton and particle distribution across a mesoscale front, and highlighted contrasted periods during the spring bloom. Finally, leveraging high frequency in situ imaging data, we investigated the fine-scale distribution and preferential position of Rhizaria, a group of understudied, fragile protists, some of which are mixotrophic. Overall, these studies demonstrate the effectiveness of in situ imaging combined with artificial intelligence to understand biophysical interactions in plankton and distribution patterns at small-scale
Green, Steven Paul. "Intelligent Person Behaviour Analysis in Low Resolution Beach Video Imagery." Thesis, Griffith University, 2011. http://hdl.handle.net/10072/366650.
Thesis (PhD Doctorate)
Doctor of Philosophy (PhD)
School of Information and Communication Technology
Science, Environment, Engineering and Technology
Full Text
Vétil, Rebeca. "Artificial Intelligence Methods to Assist the Diagnosis of Pancreatic Diseases in Radiology." Electronic Thesis or Diss., Institut polytechnique de Paris, 2024. http://www.theses.fr/2024IPPAT014.
With its increasing incidence and its five- year survival rate (9%), pancreatic cancer could be- come the third leading cause of cancer-related deaths by 2025. These figures are primarily attributed to late diagnoses, which limit therapeutic options. This the- sis aims to assist radiologists in diagnosing pancrea- tic cancer through artificial intelligence (AI) tools that would facilitate early diagnosis. Several methods have been developed. First, a method for the automatic segmentation of the pancreas on portal CT scans was developed. To deal with the specific anatomy of the pancreas, which is characterized by an elonga- ted shape and subtle extremities easily missed, the proposed method relied on local sensitivity adjust- ments using geometrical priors. Then, the thesis tack- led the detection of pancreatic lesions and main pan- creatic duct (MPD) dilatation, both crucial indicators of pancreatic cancer. The proposed method started with the segmentation of the pancreas, the lesion and the MPD. Then, quantitative features were extracted from the segmentations and leveraged to predict the presence of a lesion and the dilatation of the MPD. The method was evaluated on an external test cohort comprising hundreds of patients. Continuing towards early diagnosis, two strategies were explored to de- tect secondary signs of pancreatic cancer. The first approach leveraged large databases of healthy pan- creases to learn a normative model of healthy pan- creatic shapes, facilitating the identification of anoma- lies. To this end, volumetric segmentation masks were embedded into a common probabilistic shape space, enabling zero-shot and few-shot abnormal shape de- tection. The second approach leveraged two types of radiomics: deep learning radiomics (DLR), extracted by deep neural networks, and hand-crafted radiomics (HCR), derived from predefined formulas. The propo- sed method sought to extract non-redundant DLR that would complement the information contained in the HCR. Results showed that this method effectively de- tected four secondary signs of pancreatic cancer: ab- normal shape, atrophy, senility, and fat replacement. To develop these methods, a database of 2800 exa- minations has been created, making it one of the lar- gest for AI research on pancreatic cancer
Aberni, Yassir. "Algorithmes de reconnaissance biométrique multispectrale par l’empreinte et les veines palmaires." Electronic Thesis or Diss., Paris 8, 2021. http://www.theses.fr/2021PA080083.
Biometrics is increasingly becoming an important technology to improve security and bring solutions to services requiring authentication and/or data protection. Several biometric traits have been studied and used for biometric recognition, such as palm prints. Although many recognition methods based on the palm print have been proposed and successfully applied; where most of them generally uses only the images acquired in natural light. It is difficult to further improve the accuracy of recognition based on these palm print images due to limitations related to natural light, including the ability of identity theft attacks, degradation of palm prints over time due to several factors such as environment, ethnicity or age. Multi-spectral imaging has therefore been used to overcome these limitations. In this thesis, we propose novel biometric recognition algorithms based on physiological modalities not visible to the naked eye, called hidden, from multi-spectral imaging such as palm veins. To this end, we have developed novel biometric recognition methods using palm print and palm veins based on competitive coding using a local multiscale binary model applied on images extracted with an an ant colony optimization technique. Novel matching approaches for decision based on divergence and distance metrics have been proposed to quantify the similarity between feature images. We also proposed another novel method based on a convolutional neural network by adapting the ZFNet architecture. The experiments carried out and the comparative study with the state-of-the art, show the effectiveness of our different proposed methods for the identification and verification modes
Alves, de Lima Danilo. "Sensor-based navigation applied to intelligent electric vehicles." Thesis, Compiègne, 2015. http://www.theses.fr/2015COMP2191/document.
Autonomous navigation of car-like robots is a large domain with several techniques and applications working in cooperation. It ranges from low-level control to global navigation, passing by environment perception, robot localization, and many others in asensor-based approach. Although there are very advanced works, they still presenting problems and limitations related to the environment where the car is inserted and the sensors used. This work addresses the navigation problem of car-like robots based on low cost sensors in urban environments. For this purpose, an intelligent electric vehicle was equipped with vision cameras and other sensors to be applied in three big areas of robot navigation : the Environment Perception, Local Navigation Control, and Global Navigation Management. In the environment perception, a 2D and 3D image processing approach was proposed to segment the road area and detect the obstacles. This segmentation approach also provides some image features to local navigation control.Based on the previous detected information, a hybrid control approach for vision based navigation with obstacle avoidance was applied to road lane following. It is composed by the validation of a Visual Servoing methodology (deliberative controller) in a new Image-based Dynamic Window Approach (reactive controller). To assure the car’s global navigation, we proposed the association of the data from digital maps in order tomanage the local navigation at critical points, like road intersections. Experiments in a challenging scenario with both simulated and real experimental car show the viabilityof the proposed methodology
Rebaud, Louis. "Whole-body / total-body biomarkers in PET imaging." Electronic Thesis or Diss., université Paris-Saclay, 2024. http://www.theses.fr/2024UPAST047.
This thesis in partnership with Institut Curie and Siemens Healthineers explores the use of Positron Emission Tomography (PET) for cancer prognosis, focusing on non-Hodgkin lymphomas, especially follicular lymphoma (FL) and diffuse large B cell lymphoma (DLBCL). Assuming that current biomarkers computed in PET images overlook significant information, this work focuses on the search for new biomarkers in whole-body PET imaging. An initial manual approach validated a previously identified feature (tumor fragmentation) and explored the prognostic significance of splenic involvement in DLBCL, finding that the volume of splenic involvement does not further stratify patients with such an involvement. To overcome the empirical limitations of the manual search, a semi-automatic feature identification method was developed. It consisted in the automatic extraction of thousands of candidate biomarkers and there subsequent testing by a selection pipeline design to identify features quantifying new prognostic information. The selected biomarkers were then analysed and re-encoded in simpler and more intuitive ways. Using this approach, 22 new image-based biomarkers were identified, reflecting biological information about the tumours, but also the overall health status of the patient. Among them, 10 features were found prognostic of both FL and DLBCL patient outcome. The thesis also addresses the challenge of using these features in clinical practice, proposing the Individual Coefficient Approximation for Risk Estimation (ICARE) model. This machine learning model, designed to reduce overfitting and improve generalizability, demonstrated effectiveness in the HECKTOR 2022 challenge for predicting outcomes from head and neck cancer patients [18F]-PET/CT scans. This model was also found to overfit less than other machine learning methods on an exhaustive comparison using a benchmark of 71 medical datasets. All these developments were implemented in a software extension of a prototype developed by Siemens Healthineers
Christie, Marc. "Spécification de trajectoires de caméra sous contraintes." Nantes, 2003. http://www.theses.fr/2003NANT2116.
Brown, Marlon F. "Joint Deployable Intelligence Support System (JDISS) communications and imagery application guide for new users." Thesis, Monterey, Calif. : Springfield, Va. : Naval Postgraduate School ; Available from National Technical Information Service, 1997. http://handle.dtic.mil/100.2/ADA331705.
Thesis advisors, Gary R. Porter and Tim L. Phillips. Includes bibliographical references (p. 27). Also available online.
Sementilli, Philip Joseph Jr 1958. "Linear feature delineation in digital imagery using neural networks." Thesis, The University of Arizona, 1991. http://hdl.handle.net/10150/278012.
Guo, Xufeng. "Automated scene understanding from aerial imagery." Thesis, Queensland University of Technology, 2018. https://eprints.qut.edu.au/117973/8/Xufeng_Guo_Thesis.pdf.
Corbat, Lisa. "Fusion de segmentations complémentaires d'images médicales par Intelligence Artificielle et autres méthodes de gestion de conflits." Thesis, Bourgogne Franche-Comté, 2020. http://www.theses.fr/2020UBFCD029.
Nephroblastoma is the most common kidney tumour in children and its diagnosis is based exclusively on imaging. This work, which is the subject of our research, is part of a larger project: the European project SAIAD (Automated Segmentation of Medical Images Using Distributed Artificial Intelligence). The aim of the project is to design a platform capable of performing different automatic segmentations from source images using Artificial Intelligence (AI) methods, and thus obtain a faithful three-dimensional reconstruction. In this sense, work carried out in a previous thesis of the research team led to the creation of a segmentation platform. It allows the segmentation of several structures individually, by methods such as Deep Learning, and more particularly Convolutional Neural Networks (CNNs), as well as Case Based Reasoning (CBR). However, it is then necessary to automatically fuse the segmentations of these different structures in order to obtain a complete relevant segmentation. When aggregating these structures, contradictory pixels may appear. These conflicts can be resolved by various methods based or not on AI and are the subject of our research. First, we propose a fusion approach not focused on AI using the combination of six different methods, based on different imaging and segmentation criteria. In parallel, two other fusion methods are proposed using, a CNN coupled to the CBR for one, and a CNN using a specific existing segmentation learning method for the other. These different approaches were tested on a set of 14 nephroblastoma patients and demonstrated their effectiveness in resolving conflicting pixels and their ability to improve the resulting segmentations
Wallis, David. "A study of machine learning and deep learning methods and their application to medical imaging." Thesis, université Paris-Saclay, 2021. http://www.theses.fr/2021UPAST057.
We first use Convolutional Neural Networks (CNNs) to automate mediastinal lymph node detection using FDG-PET/CT scans. We build a fully automated model to go directly from whole-body FDG-PET/CT scans to node localisation. The results show a comparable performance to an experienced physician. In the second half of the thesis we experimentally test the performance, interpretability, and stability of radiomic and CNN models on three datasets (2D brain MRI scans, 3D CT lung scans, 3D FDG-PET/CT mediastinal scans). We compare how the models improve as more data is available and examine whether there are patterns common to the different problems. We question whether current methods for model interpretation are satisfactory. We also investigate how precise segmentation affects the performance of the models. We first use Convolutional Neural Networks (CNNs) to automate mediastinal lymph node detection using FDG-PET/CT scans. We build a fully automated model to go directly from whole-body FDG-PET/CT scans to node localisation. The results show a comparable performance to an experienced physician. In the second half of the thesis we experimentally test the performance, interpretability, and stability of radiomic and CNN models on three datasets (2D brain MRI scans, 3D CT lung scans, 3D FDG-PET/CT mediastinal scans). We compare how the models improve as more data is available and examine whether there are patterns common to the different problems. We question whether current methods for model interpretation are satisfactory. We also investigate how precise segmentation affects the performance of the models
Toony, Zahra. "Extracting structured models from raw scans of manufactured objects : a step towards embedded intelligent handheld 3D scanning." Doctoral thesis, Université Laval, 2015. http://hdl.handle.net/20.500.11794/26270.
The availability of fast and accurate 3D sensors has favored the development of different applications in assembly, inspection, Computer-Aided Design (CAD), reverse engineering, mechanical engineering, medicine, and entertainment, to list just a few. While 2D cameras capture 2D images of the surface of objects, either black-and-white or color, 3D cameras provide information on the geometry of an object surface. Today, newly introduced 3D cameras can acquire the appearance and geometry of objects concurrently. The popularity and availability of such these 3D models has opened new fields of interests, such as 3D model segmentation, 3D model recognition, estimation of 3D models’ parameters and even 3D modelling. In this project, we aim at recognizing different parts (primitives) of a 3D object, intelligently. For this purpose, we first prepare a database of 3D CAD primitives (i.g. planes, cylinders, cones, spheres and, tori). Then using segmentation algorithms, the complex objects are decomposed into their primitives and, by utilizing recognition techniques, a descriptor is extracted and associated to each primitive and, finally, a classifier is trained to learn the properties of primitives. The manuscript investigates different methods related to these challenges. An additional step is also proposed in this project which estimates the parameters of primitives and generate the CAD primitives that completes the whole process of reverse engineering.
Ponce, Jean. "Représentation des objets tridimensionnels." Paris 11, 1988. http://www.theses.fr/1988PA112369.
This thesis is concerned with the geometric representation of three-dimensional objects in the context of computer vision. The thesis is divided into three parts: generalized cylinders and their images, CAD (Computer Aided Design) models, range images measured with a laser rangefinder. In the first part, the differential geometry of straight homogeneous generalized cylinders (SHGC's) is studied. The results obtained are used to prove several new uniqueness results, to characterize the silhouette of SHGC's, and to prove several properties of the contours of these objects which are independent from the viewing direction. These properties are used in turn for image segmentation. A CAD system is described in the second part. Complex solids are obtained by combining SHGC's through set operations. A hierarchical data structure is used to compute the set operations in an efficient and robust manner. The system described includes fast and powerful rendering algorithms and tools for building and modifying easily primitives and complex objects. In the third part, range images are represented at different scales by their principal curvatures and directions. An analytical representation of range discontinuities is used to segment these images. A graph of smooth surfaces separated by these discontinuities is built, and these surfaces are described by their lines of curvature, by planar and spherical regions, and by certain types of generalized cylinders
Thomas, Vincent. "Modélisation 3D pour la réalité augmentée : une première expérimentation avec un téléphone intelligent." Thesis, Université Laval, 2011. http://www.theses.ulaval.ca/2011/27904/27904.pdf.
Recently, a new genre of software applications has emerged allowing the general public to browse their immediate environment using their smartphone: Mobile Augmented Reality (MAR) applications. The growing popularity of this type of application is triggered by the fast evolution of smartphones. These ergonomic mobile platforms embed several pieces of equipment useful to deploy MAR (i.e. digital camera, GPS receiver, accelerometers, digital compass and now gyroscope). In order to achieve a strong augmentation of the reality in terms of user’s immersion and interactions, a 3D model of the real environment is generally required. The 3D model can be used for three different purposes in these MAR applications: 1) to manage the occlusions between real and virtual objects; 2) to provide accurate camera pose (position/orientation) calculation; 3) to support the augmentation and interactions. However, the availability of such 3D models is limited and therefore preventing MAR application to be used anywhere at anytime. In order to overcome such constraints, this proposed research thesis is aimed at devising a new approach adapted to the specific context of MAR applications and dedicated to the simple and fast production of 3D models. This approach was implemented on the iPhone 3G platform and evaluated according to precision, rapidity, simplicity and efficiency criteria. Results of the evaluation underlined the capacity of the proposed approach to provide, in about 3 minutes, a simple 3D model of a building using smartphone while achieving accuracy of 5 meters and higher.
Duran, Audrey. "Intelligence artificielle pour la caractérisation du cancer de la prostate par agressivité en IRM multiparamétrique." Thesis, Lyon, 2022. http://theses.insa-lyon.fr/publication/2022LYSEI008/these.pdf.
Prostate cancer (PCa) is the most frequently diagnosed cancer in men in more than half the countries in the world and the fifth leading cause of cancer death among men in 2020. Diagnosis of PCa includes multiparametric magnetic resonance imaging acquisition (mp-MRI) - which combines T2 weighted (T2-w), diffusion weighted imaging (DWI) and dynamic contrast enhanced (DCE) sequences - prior to any biopsy. The joint analysis of these multimodal images is time demanding and challenging, especially when individual MR sequences yield conflicting findings. In addition, the sensitivity of MRI is low for less aggressive cancers and inter-reader reproducibility remains moderate at best. Moreover, visual analysis does not currently allow to determine the cancer aggressiveness, characterized by the Gleason score (GS). This is why computer-aided diagnosis (CAD) systems based on statistical learning models have been proposed in recent years, to assist radiologists in their diagnostic task, but the vast majority of these models focus on the binary detection of clinically significant (CS) lesions. The objective of this thesis is to develop a CAD system to detect and segment PCa on mp-MRI images but also to characterize their aggressiveness, by predicting the associated GS. In a first part, we present a supervised CAD system to segment PCa by aggressiveness from T2-w and ADC maps. This end-to-end multi-class neural network jointly segments the prostate gland and cancer lesions with GS group grading. The model was trained and validated with a 5-fold cross-validation on a heterogeneous series of 219 MRI exams acquired on three different scanners prior prostatectomy. Regarding the automatic GS group grading, Cohen’s quadratic weighted kappa coefficient (κ) is 0.418 ± 0.138, which is the best reported lesion-wise kappa for GS segmentation to our knowledge. The model has also encouraging generalization capacities on the PROSTATEx-2 public dataset. In a second part, we focus on a weakly supervised model that allows the inclusion of partly annotated data, where the lesions are identified by points only, for a consequent saving of time and the inclusion of biopsy-based databases. Regarding the automatic GS group grading on our private dataset, we show that we can approach performance achieved with the baseline fully supervised model while considering 6% of annotated voxels only for training. In the last part, we study the contribution of DCE MRI, a sequence often omitted as input to deep models, for the detection and characterization of PCa. We evaluate several ways to encode the perfusion from the DCE MRI information in a U-Net like architecture. Parametric maps derived from DCE MR exams are shown to positively impact segmentation and grading performance of PCa lesions
Dalsasso, Emanuele. "Deep learning for SAR imagery : from denoising to scene understanding." Electronic Thesis or Diss., Institut polytechnique de Paris, 2022. http://www.theses.fr/2022IPPAT008.
Synthetic Aperture Radars (SARs) can collect data for Earth Observation purposes regardless of the daylight or cloud cover. Nowadays, thanks to the Copernicus program of the European Space Agency, a huge amount of SAR data is freely available. However, the exploitation of satellite SAR images is limited by the presence of strong fluctuations in the backscattered signal. Indeed, SAR images are corrupted by speckle, a phenomenon inherent to coherent imaging systems. In this Ph.D thesis, we aim to improve the interpretation of SAR images by resorting to speckle reduction techniques. Existing approaches are based on Goodman’s model, which describes the speckle component as a spatially uncorrelated multiplicative noise. In the computer vision field, denoising methods relying on Convolutional Neural Networks (deep learning approaches) have led to great improvements and provide nowadays state-of-the-art results. We propose to use deep learning-based denoising techniques to reduce speckle from SAR images (despeckling methods). At first, we study the adaptation of supervised techniques that minimize a certain distance between the estimation provided by the CNN and a reference image, also called “groundtruth”. We propose to create a dataset of reference images by averaging multi-temporal images acquired over the same area. Pairs of reference and corrupted images can be generated by synthetizing speckle following Goodman’s model. However, in real images the speckle component is spatially correlated which typically requires subsampling these images by a factor 2 to reduce the spatial correlations, which also degrades the spatial resolution. Given the limits of supervised approaches and inspired by noise2noise, a self-supervised denoising method, we propose to train our networks directly on actual SAR images. The principle of self-supervised denoising methods is the following: if a signal contains a deterministic component and a random component, then a network trained to predict a new signal realization from a first independent signal realization will only predict the deterministic component, i.e., the underlying scene, thereby suppressing the speckle. In the method we have developed, SAR2SAR, we leverage multi-temporal SAR series to obtain independent realizations of the same scene, under the hypothesis of temporally decorrelated speckle. Changes are compensated by devising an iterative training strategy. SAR2SAR is thus trained directly on images with spatially correlated speckle and can readily be applied on SAR images without subsampling, providing high-quality results. The training of SAR2SAR is quite heavy: it is articulated in several steps to compensate changes and a dataset comprising stacks of images must be built. With our approach “MERLIN”, we alleviate the training by proposing a single-image learning strategy. Indeed, in single-look-complex SAR images, real and imaginary parts are mutually independent and can benaturally exploited to train CNNs with self-supervision. We show the potential of this training framework for three imaging modalities, different in terms of spatial resolution, textures, and speckle spatial correlation. For the sake of open science, the code associated to each algorithm developed is made freely available
Charest, Jonathan. "Système intelligent de détection et diagnostic de fautes en tomographie d'émission par positrons." Thèse, Université de Sherbrooke, 2017. http://hdl.handle.net/11143/11628.
Workman, Scott. "Leveraging Overhead Imagery for Localization, Mapping, and Understanding." UKnowledge, 2018. https://uknowledge.uky.edu/cs_etds/64.
Richard, Jean-Alexandre. "Synthèse de sondes luminescentes utilisant un bras réactif auto-immolable : application à la détection de peptidases." Thesis, Rouen, INSA, 2008. http://www.theses.fr/2008ISAM0011.
Optical imaging is currently revolutionizing pre-clinic diagnosis and drug development. In that context, QUIDD develops smart probes able to follow biological events involved in biological disorders. The aim of this PhD work was to provide a general method for the synthesis of luminescent probes able to detect proteases. For that purpose, probes composed of a peptide substrate and a phenolic luminescent moiety connected by a self-immolative linker were developed. Two strategies were investigated: a first strategy involved phenolic pro-fluorophores, whose fluorescence is quenched when their phenol functionality is substituted. A second strategy took advantage of 1,2-dioxetanes whose liberation in the medium results in a spontaneous light emission. The first objective of this work was to provide a proof of concept of these strategies, especially for the detection of caspase-3, a key enzyme involved in the apoptotic process. The second part of this work was devoted to the extension of the strategy to in vivo imaging
Chambe, Mathieu. "Improving image quality using high dynamic range and aesthetics assessment." Electronic Thesis or Diss., Université de Rennes (2023-....), 2023. http://www.theses.fr/2023URENS015.
To cope with the increasing amount of visual content available, it is important to devise automatic processes that can sort, improve, compress or store images and videos. In this thesis, we propose two different approaches to software-based image improvement. First, we propose a study on existing aesthetics assessment algorithms. These algorithms are based on supervised neural networks. We have collected several datasets of images, and we have tested different models using these images. We report here the performances of such networks, as well as an idea to improve the already trained networks. Our study shows that the features needed to accurately predict the aesthetics of competitive and professional are different but can be learned simultaneously by a single network. In a second time, we propose to work with High Dynamic Range (HDR) images. We present here a new operator to increase the dynamic range of images called HDR-LFNet, that merges the output of existing operators and therefore, consists in far fewer parameters. Besides, we evaluate our method through objective metrics and a user study. We show that our method is on-par with the state-of-the-art according to objective metrics, but is preferred by observers during the user study, while using less resources overall
Khatir, Leyla. "Recherche d'algorithmes de localisation de routes dans les images haute resolution : application a l'imagerie spot." Toulouse 3, 1988. http://www.theses.fr/1988TOU30080.
Moussa, Richard. "SEGMENTATION MULTI-AGENTS EN IMAGERIE BIOLOGIQUE ET MÉDICALE : APPLICATION AUX IRM 3D." Phd thesis, Université Sciences et Technologies - Bordeaux I, 2011. http://tel.archives-ouvertes.fr/tel-00652445.
Grandclaude, Virgile. "Synthèse de sondes chémiluminescentes et profluorescentes pour des applications en imagerie in vivo." Thesis, Rouen, INSA, 2011. http://www.theses.fr/2011ISAM0009.
Optical molecular imaging is now playing a pivotal role both in pre-clinical diagnosis and drug development. Indeed, this is a valuable tool for the real time detection and monitoring of living cells either through the use of structurally simple labels or more recently by means of sophisticated fluorescent probes, called “smart” probes and only activatable upon specific interaction with the targeted bio-analyte. The aim of this PhD work was the design of new synthetic tools aimed at optimizing physico-chemical and optical properties of fluorescent probes intended for challenging in vivo imaging applications. We have focused on the pro-fluorescence and chemiluminescence approaches. New phenol-based pro-fluorophores have been developed by using an original bis-coumarinic scaffold. In the context of the chemistry of fluorophores, we have also investigated a general method for the water-solubilisation of phenol-based fluorophore belonging to the coumarin and xanthene families. Our research in chemiluminescence has led the synthesis of new chemiluminophores covalently linked to fluorescent organic dyes aimed at increasing the emission efficiency in the red region of such chemiluminophores. Thus, the first chemiluminescent “energy transfer cassettes” based on a 1,2-dioxetane scaffold have been obtained
Dercle, Laurent. "Radiomics : Artificial Intelligence Driven Extraction of Information from Medical Images to Guide Clinical Decision in Cancer Patients Treated with Immunotherapy." Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLS435.
Immunotherapies targeting the programmed cell death receptor-1 and ligand-1 pathways (anti-PD(L)1) have emerged as an effective treatment for a variety of cancers. Anti-PD(L)1 is a paradigm shift in the treatment of cancers since its activity relies on restoring an efficient anti-tumor T-cell response. Two main reasons explain the need to investigate biomarkers forecasting survival and predicting the anti-cancer efficacy of anti-PD(L)1. First, an excess of death has been observed in the experimental arm of randomized phase III trials comparing anti-PD(L)1 immunotherapies to chemotherapy for multiple cancers. Among the controversial hypotheses that would explain this observation are frequently mentioned the lack of effectiveness of anti-PD(L)1 in patients with a fast-growing disease (so-called "fast progressors") vs. a paradoxical effect of disease acceleration under immunotherapy (so-called "hyperprogressors"). Second, imaging response criteria play a pivotal role in guiding cancer patient management and define a "wait and see strategy" for patients treated with anti-PD(L)1 in monotherapy with progressive disease. The distinct mechanisms of anti-PD(L)1, which restore the immune system's anti-tumor capacity, leads to unconventional immune-related phenomena. From a medical imaging standpoint, it translates into pseudoprogression, hyperprogression, abscopal effect, and immune-related adverse events. We leveraged machine learning approaches to challenge the prognostic/predictive factors and identify which imaging biomarkers are associated with early death upon anti-PD(L)1 immunotherapy. We mined transcriptomic data to determine the biological pathways related to these prognostic/predictive factors. Our results demonstrate that a limited subset of imaging biomarkers can forecast overall survival. The classification of these imaging biomarkers into distinct hallmarks provides a conceptual structure and logical coherence delineating the interconnections between them. These hallmarks can be understood as distinct physiological circuits disrupted by cancer that are linked to shorter survival: liver organotropism, high tumor burden, high heterogeneity in tumor vascularity or metabolism, infiltration along tumor boundaries, irregularity in tumor shape, high glucose consumption, sarcopenia, and high bone marrow metabolism. Using machine-learning, we demonstrated that increased baseline serum lactate dehydrogenase and the presence of liver metastasis on CT-scan are two independent drivers of premature death after anti-PD(L)1 initiation. Transcriptomic analysis identified actionable pathways amenable to novel treatments, which could improve anti-PD(L)1 efficacy. From a broader perspective, this demonstrates the need to continue to develop advanced imaging technology to enhance the monitoring of cancer patients treated with immunotherapeutics. This involves analyzing and linking data in pathology, oncology, and radiology, and the ability to work with extensive datasets. Therefore, there is a need to develop comprehensive programs of radiomics for predictive tools that benefit diagnosis, assessment, and management of all types of cancer patients. In conclusion, radiomics driven precision medicine approaches could improve the lives of cancer patients treated with cancer immunotherapy
Galibourg, Antoine. "Estimation de l'âge dentaire chez le sujet vivant : application des méthodes d'apprentissage machine chez les enfants et les jeunes adultes." Electronic Thesis or Diss., Toulouse 3, 2022. http://thesesups.ups-tlse.fr/5355/.
Statement of the problem: In the living individual, the estimation of dental age is a parameter used in orthopedics or dentofacial orthodontics or in pediatrics to locate the individual on its growth curve. In forensic medicine, the estimation of dental age allows to infer the chronological age for a regression or a classification task. There are physical and radiological methods. While the latter are more accurate, there is no universal method. Demirjian created the most widely used radiological method almost 50 years ago, but it is criticized for its accuracy and for using reference tables based on a French-Canadian population sample. Objective: Artificial intelligence, and more particularly machine learning, has allowed the development of various tools with a learning capacity on an annotated database. The objective of this thesis was to compare the performance of different machine learning algorithms first against two classical methods of dental age estimation, and then between them by adding additional predictors. Material and method: In a first part, the different methods of dental age estimation on living children and young adults are presented. The limitations of these methods are exposed and the possibilities to address them with the use of machine learning are proposed. Using a database of 3605 panoramic radiographs of individuals aged 2 to 24 years (1734 girls and 1871 boys), different machine learning methods were tested to estimate dental age. The accuracies of these methods were compared with each other and with two classical methods by Demirjian and Willems. This work resulted in an article published in the International Journal of Legal Medicine. In a second part, the different machine learning methods are described and discussed. Then, the results obtained in the article are put in perspective with the publications on the subject in 2021. Finally, a perspective of the results of the machine learning methods in relation to their use in dental age estimation is made. Results: The results show that all machine learning methods have better accuracy than the conventional methods tested for dental age estimation under the conditions of their use. They also show that the use of the maturation stage of third molars over an extended range of use to 24 years does not allow the estimation of dental age for a legal issue. Conclusion: Machine learning methods fit into the overall process of automating dental age determination. The specific part of deep learning seems interesting to investigate for dental age classification tasks
Chauvin, Thomas. "Développement d'agents de contraste intelligents pour l'Imagerie par Résonance Magnétique (IRM)." Thesis, Orléans, 2010. http://www.theses.fr/2010ORLE2011.
Today, Magnetic Resonance Imaging is one of the most powerful diagnostic techniques in the clinics. Amongfuture perspectives, molecular imaging applications based on smart contrast agents which are responsive to various physico-chemical parameters, are particularly attractive. In this work, we present the synthesis and physico-chemical characterisation of novel lanthanide complexes with the aim of developing smart contrast agents for the detection of enzyme activity or calcium concentration.The complexes designed to give an MRI response to an enzyme are based on the original concept of coupling an enzyme-specific substrate to a macrocyclic LnIII chelate via a self-immolative linker. The structural changes following enzymatic cleavage of the substrate and destruction of the self-immolative armare expected to induce variation of the relaxivity or the CEST properties of the LnIII complexes. Though we failed creating GdIII agents with a T1 response upon enzymatic reaction, several YbIII or EuIII complexes were synthesized that provide an important change in their CEST properties. Some of them, bearing a pyridine-derivative arm which is an efficient sensitizer of lanthanide luminescence, act also as enzyme-responsive NIR or visible emitting optical probes.We have developed a Ca-responsive agent combining a DOTA-tetraamide LnIII chelator with animinodiacetate unit for calcium coordination. The EuIII and YbIII complexes show an important decrease in the CEST effect in response to Ca2+. The parallel application of the two complexes allows for ratiometric approaches where the detected MRI response is independent of the concentration of the agent
Kubler, Samuel. "Statistical methods for the robust extraction of objects’ spatio-temporal relations in bioimaging – Application to the functional analysis of neuronal networks in vivo." Electronic Thesis or Diss., Sorbonne université, 2023. http://www.theses.fr/2023SORUS455.
The neural code, i.e. how interconnected neurons can perform complex operations, allowing the quick adaptation of animals to their environment, remains an open question and an intensive field of research both in experimental and computational neurosciences. Advances in molecular biology and microscopy have recently made it possible to monitor the activity of individual neurons in living animals and, in the case of small animals containing only a few thousands of neurons, to measure the activity of the entire nervous system. However, the mathematical framework that would bridge the gap between single neuron activity and the emergent computational properties of neuronal ensembles is missing.In the thesis manuscript, we introduce a sequential statistical processing pipeline that efficiently and robustly extracts neuronal ensembles from calcium imagery of neuronal activity. In particular, we develop a Bayesian inference framework based on a biologically interpretable model to extract neuronal ensembles characterized by noise, asynchrony and overlapping. The provided tool demonstrates that a Gibbs sampling routine can efficiently estimate statistical parameters and hidden variables to uncover neuronal ensembles based on synchronization patterns both on synthetic data and on various experimental datasets from mice and zebrafish visual cortex to Hydra Vulgaris. The thesis equally develops a point process statistical framework to quantify how neuronal ensembles encode evoked stimuli or spontaneous behaviors in living animals. This versatile tool is also used for the inference of the functional connectivity of neuronal activity or the automatically calibration procedure of the spike inference algorithms applied to calcium recordings. For the providing algorithms to be largely spread in the neurobiologist community, results are supported by interpretable biological estimates, statistical evidence, rigorous mathematical proofs, and free-available software. Our contributive implementation, that goes from pixel intensity to estimated neuronal ensembles, equally identify from the synchronous firing patterns of neuronal ensembles, neurons with specific roles that can be used to predict, improve, or alter the behaviors of living animals. The provided framework unravels the emergence of collective properties from the recording of extremely varying individual signals that make the neural code still elusive
Dong, Min. "Development of an intelligent recommendation system to garment designers for designing new personalized products." Thesis, Lille 1, 2017. http://www.theses.fr/2017LIL10025/document.
In my PhD research project, we originally propose a Designer-oriented Intelligent Recommendation System (DIRS) for supporting the design of new personalized garment products. For developing this system, we first identify the key components of a garment design process, and then set up a number of relevant databases, from which each design scheme can be formed. Second, we acquire the anthropometric data and designer’s perception on body shapes by using a 3D body scanning system and a sensory evaluation procedure. Third, an instrumental experiment is conducted for measuring the technical parameters of fabrics, and five sensory experiments are carried out in order to acquire designers’ knowledge. The acquired data are used to classify body shapes and model the relations between human bodies and the design factors. From these models, we set up an ontology-based design knowledge base. This knowledge base can be updated by dynamically learning from new design cases. On this basis, we put forward the knowledge-based recommendation system. This system is used with a newly developed design process. This process can be performed repeatedly until the designer’s satisfaction. The proposed recommendation system has been validated through a number of successful real design cases
Richard, Jean-Alexandre. "Synthèse de sondes luminescentes utilisant un bras réactif auto-immolable : application à la détection de peptidases." Electronic Thesis or Diss., Rouen, INSA, 2008. http://www.theses.fr/2008ISAM0011.
Optical imaging is currently revolutionizing pre-clinic diagnosis and drug development. In that context, QUIDD develops smart probes able to follow biological events involved in biological disorders. The aim of this PhD work was to provide a general method for the synthesis of luminescent probes able to detect proteases. For that purpose, probes composed of a peptide substrate and a phenolic luminescent moiety connected by a self-immolative linker were developed. Two strategies were investigated: a first strategy involved phenolic pro-fluorophores, whose fluorescence is quenched when their phenol functionality is substituted. A second strategy took advantage of 1,2-dioxetanes whose liberation in the medium results in a spontaneous light emission. The first objective of this work was to provide a proof of concept of these strategies, especially for the detection of caspase-3, a key enzyme involved in the apoptotic process. The second part of this work was devoted to the extension of the strategy to in vivo imaging
La, Barbera Giammarco. "Learning anatomical digital twins in pediatric 3D imaging for renal cancer surgery." Electronic Thesis or Diss., Institut polytechnique de Paris, 2022. http://www.theses.fr/2022IPPAT040.
Pediatric renal cancers account for 9% of pediatric cancers with a 9/10 survival rate at the expense of the loss of a kidney. Nephron-sparing surgery (NSS, partial removal of the kidney) is possible if the cancer meets specific criteria (regarding volume, location and extent of the lesion). Indication for NSS is relying on preoperative imaging, in particular X-ray Computerized Tomography (CT). While assessing all criteria in 2D images is not always easy nor even feasible, 3D patient-specific models offer a promising solution. Building 3D models of the renal tumor anatomy based on segmentation is widely developed in adults but not in children. There is a need of dedicated image processing methods for pediatric patients due to the specificities of the images with respect to adults and to heterogeneity in pose and size of the structures (subjects going from few days of age to 16 years). Moreover, in CT images, injection of contrast agent (contrast-enhanced CT, ceCT) is often used to facilitate the identification of the interface between different tissues and structures but this might lead to heterogeneity in contrast and brightness of some anatomical structures, even among patients of the same medical database (i.e., same acquisition procedure). This can complicate the following analyses, such as segmentation. The first objective of this thesis is to perform organ/tumor segmentation from abdominal-visceral ceCT images. An individual 3D patient model is then derived. Transfer learning approaches (from adult data to children images) are proposed to improve state-of-the-art performances. The first question we want to answer is if such methods are feasible, despite the obvious structural difference between the datasets, thanks to geometric domain adaptation. A second question is if the standard techniques of data augmentation can be replaced by data homogenization techniques using Spatial Transformer Networks (STN), improving training time, memory requirement and performances. In order to deal with variability in contrast medium diffusion, a second objective is to perform a cross-domain CT image translation from ceCT to contrast-free CT (CT) and vice-versa, using Cycle Generative Adversarial Network (CycleGAN). In fact, the combined use of ceCT and CT images can improve the segmentation performances on certain anatomical structures in ceCT, but at the cost of a double radiation exposure. To limit the radiation dose, generative models could be used to synthesize one modality, instead of acquiring it. We present an extension of CycleGAN to generate such images, from unpaired databases. Anatomical constraints are introduced by automatically selecting the region of interest and by using the score of a Self-Supervised Body Regressor, improving the selection of anatomically-paired images between the two domains (CT and ceCT) and enforcing anatomical consistency. A third objective of this work is to complete the 3D model of patient affected by renal tumor including also arteries, veins and collecting system (i.e. ureters). An extensive study and benchmarking of the literature on anatomic tubular structure segmentation is presented. Modifications to state-of-the-art methods for our specific application are also proposed. Moreover, we present for the first time the use of the so-called vesselness function as loss function for training a segmentation network. We demonstrate that combining eigenvalue information with structural and voxel-wise information of other loss functions results in an improvement in performance. Eventually, a tool developed for using the proposed methods in a real clinical setting is shown as well as a clinical study to further evaluate the benefits of using 3D models in pre-operative planning. The intent of this research is to demonstrate through a retrospective evaluation of experts how criteria for NSS are more likely to be found in 3D compared to 2D images. This study is still ongoing
Li, Songlin. "Modélisation d'un implant médical intelligent dans son environnement pour le monitorage fonctionnel de la moelle épinière." Electronic Thesis or Diss., Sorbonne université, 2022. http://www.theses.fr/2022SORUS478.
The spinal cord (SC) is part of the central nervous system. It originates in the brain and is responsible for important functions, such as the transmission of nerve information between the brain and the periphery, as well as reflexes in the trunk and extremities. Trauma to the SC can result in loss of sensation and control of a body part in minor cases, or total paralysis in severe cases. Real-time monitoring of the functional status of the ME during spinal surgery, such as scoliosis surgery, is essential to avoid the serious consequences of unintentional damage to the SC during the surgical procedure. However, there is currently no method to obtain sufficient information about the changes in the function of the SC, before it is affected. The use of multi-wavelength Diffuse Optical Imaging (DOI) seems to be a promising alternative to meet this need. Indeed, it allows real-time monitoring of the hemodynamic characteristics of the SC, which are important indicators for assessing its functional status. In this thesis, we propose to implement this technique through the modeling, realization and testing of a specific device. First of all, the modeling of the monitoring system in interaction with the biological environment of the SC is carried out in SystemC and SystemC-AMS. This allows us to estimate the performances of the system and to optimize them, but also to study the main physiological characteristics of the SC. This multi-domain modeling approach (e.g., optics, biology, electronics) has the advantages of being flexible, easy to modify, and adaptable to any other type of application. Secondly, the prototyping of the monitoring system is realized. Based on the results of the simulation and the experimental data, a wireless transmission module of the last generation is implemented to gain in ergonomics compared to the targeted application. Finally, data from the experiments on the porcine model are processed to extract relevant physiological information, which provides a reference for a signal processing process of the SC. These three parts complement each other to lead to a proposal for a medical device that could be used in the future on human patients to help doctors determine the functional status of the SC in real-time and avoid some irreversible trauma during surgery of the spine and its periphery
Oukhatar, Fatima. "Design, synthesis and characterization of neurotransmitter responsive probes for magnetic resonance and optical imaging." Thesis, Orléans, 2012. http://www.theses.fr/2012ORLE2076/document.
In spite of the key role of neurotransmitters (NTs) in signal transduction, their non-invasive in vivo monitoring remains an important challenge. Magnetic resonance imaging (MRI) has recently been demonstrated as a promising technique to non-invasively visualize physiological events with excellent temporal and spatial resolution. In particular, smart MRI contrast agents that are able to report on the physico-chemical status of the tissues, start to have a strong impact in neuroscience. The objective of this work was the design, synthesis and in vitro characterization of a series of lanthanide-based probes responsive to NTs with the aim to track in vivo concentration changes of NTs using MR or optical imaging. The design of our imaging probes relies on a dual binding approach of zwitterionic NTs to the Ln3+ complexes, involving interactions (i) between a positively charged Ln3+ chelate and the carboxylate function of the NTs and (ii) between an azacrown ether appended on the chelate and the amine group of the neurotransmitters. Some of the novel contrast agents were found to exhibit high relaxivities and a remarkable relaxivity response towards NTs, though little selectivity against bicarbonate. In order to apply a bimodal MRI/optical imaging approach, we have also incorporated a benzophenone moiety into the chelate to sensitize the near-infrared emitting Ln3+ ions. The Yb3+ analogue proved to be highly sensitive to NTs
Pourchot, Aloïs. "Improving Radiographic Diagnosis with Deep Learning in Clinical Settings." Electronic Thesis or Diss., Sorbonne université, 2022. http://www.theses.fr/2022SORUS421.
The impressive successes of deep learning over the course of the past decade have reinforced its establishment as the standard modus operandi to solve difficult machine learning problems, as well as enabled its swift spread to manifold domains of application. One such domain, which is at the heart of this PhD, is medical imaging. Deep learning has made the thrilling perspective of relieving medical experts from a fraction of their burden through automated diagnosis a reality. Over the course of this thesis, we were led to consider two medical problems: the task of fracture detection, and the task of bone age assessment. For both of them, we strove to explore possibilities to improve deep learning tools aimed at facilitating their diagnosis. With this objective in mind, we have explored two different strategies. The first one, ambitious yet arrogant, has led us to investigate the paradigm of neural architecture search, a logical succession to deep learning which aims at learning the very structure of the neural network model used to solve a task. In a second, bleaker but wiser strategy, we have tried to improve a model through the meticulous analysis of the data sources at hands. In both scenarios, a particular care was given to the clinical relevance of our different results and contributions, as we believed that the practical anchoring of our different contrivances was just as important as their theoretical design
Wassermann, Demian. "Automated in vivo dissection of white matter structures from diffusion magnetic resonance imaging." Nice, 2010. http://www.theses.fr/2010NICE4066.
The brain is organized in networks that are made up of tracks connecting different regions. These networks are important for the development of brain functions such as language. Lesions and cognitive disorders are sometimes better explained by disconnection mechanisms between cerebral regions than by damage of those regions. Despite several decades of tracing these networks in the brain, our knowledge of cerebral connections has progressed very little since the beginning of the last century. Recently, we have seen a spectacular development of magnetic resonance imaging (MRI) techniques for the study of the living human brain. One technique for exploring white matter (WM) tissue characteristics and pathway in vivo is diffusion MRI (dMRI). Particulary, dMRI tractography facilitates the tracing the WM tracts in vivo. DMRI is a promising technique to explore the anatomical basis of human cognition and its disorders. The motivation of this thesis is the in vivo dissection of the WM. This procedure isolates the WM tracts that play a role in a particular function or disorder of the brain so they can be analysed. Manually performing this task requires a great knowledge of brain anatomy and several hours of work. Hence, the development of a technique to automatically perform the identification of WM structures is of utmost importance. This thesis has several contributions : we develop means for the automatic dissection of WM tracts from dMRI, this is based on a mathematical framework for the WM and its tracts ; using these tools, we develop techniques to analyse the spinal chord and to find group differences in the WM particulary between healthy and schizophrenic subjects
Vassilo, Kyle. "Single Image Super Resolution with Infrared Imagery and Multi-Step Reinforcement Learning." University of Dayton / OhioLINK, 2020. http://rave.ohiolink.edu/etdc/view?acc_num=dayton1606146042238906.
Antoine, Mathieu. "Les techniques algorithmiques du codage Octree appliquées à l'analyse de volumes tridimensionnels de données numériques." Grenoble 1, 1988. http://www.theses.fr/1988GRE10042.
Depersin-Venon, Yvonne. "Relations entre image mentale et pensée opératoire : une étude comparative entre un échantillon d'enfants de classes primaires et un échantillon d'enfants en difficulté scolaire." Paris 5, 1990. http://www.theses.fr/1990PA05H042.
The piagetion hypothesis about mental imagery asserts that the development of imaginal representation of transformations abilities is subordinated to "mathematico-logical" operations. This hypothesis has been tested by comparing two groups of children. In one of thems children are mentally deficient, not in the other. If the piagetian hypothesis were true, then a retard in the "mathematico-logical" operations development should imply a developmental retard in the mental imagery. Any other kind of relation would disprove the subordination hypothesis. Two groups of 42 children from 9 to 12 have been examined over seven tasks : operational imagery and imitation tasks. One of these groups is composed of primary school children (from third to fifth grade), the other group includes children from special education. The comparative analysis findings din't disprove the piagetian hypthesis : results show an homogeneous retard in the mentally deficient children abilities in comparison with the primary school children. This finding is consistent with other hypothesis which are discussed. About the mental imagery origin, results don't show a relation between imitation an imagery as asserted in the piagetian hypothesis. The analysis of the relations structure between the different tasks, by the regression analysis, show that this structure is not the same in the two groups of children. It seems that mentally deficient children use imagery as a mode of processing in every kind of task, even in the "logical" tasks. The results extensions are discussed in conclusion
Ambroise, Corentin. "Structure-aware neural networks in the study of multi-modal population cohorts : an application to mental health." Electronic Thesis or Diss., université Paris-Saclay, 2024. http://www.theses.fr/2024UPAST065.
It is currently acknowledged that relying solely on conventional classification strategies from a single data source is not effective to understand, diagnose or prognose psychiatric syndromes. The classification targets simply rely on clinician labels that alone do not express a very large variability. In 2009, the Research Domain Criteria (RDoC) recommended a more comprehensive approach to study psychiatric disorders by incorporating diverse data types that cover various levels of life organization (e.g., imaging, genetic, symptoms). The proposal suggests that a thorough description of a pathology requires consideration of multiple dimensions, which may be shared across different psychiatric syndromes and even contribute to non-pathological variability. Efficient frameworks for unsupervised learning, specifically designed for multivariate and multimodal approaches, are anticipated to offer methodologies for handling and integrating the kind of datasets advocated by the RDoC. Deep learning allows us to learn in multimodal settings with modality-specific structure and intermodality correlation structure.To model intra-modality structure, we use specific convolutional neural networks that enable to learn from cortical brain measures distributed across a spherical mesh and thus reveal original biomarkers. In this context, we propose 5 data augmentations and apply them in one of the many novel self-supervised learning schemes relying moslty on data augmentation. This work allows deep representation learning to properly initialize network on huge healthy patient cohorts and then transfer them to study clinical pathology of interest in smaller cohorts.On the other hand, we have identified multi-view variational auto encoders as good candidates to integrate multiple modalities. Moreover, we challenge the common assumption that neural networks are not interpretable. We use a digital avatar procedure as an interpretability module capable of reporting the inter-view relationships learned within a multi-view autoencoder. In particular, we integrate this procedure into a novel framework that combines multiple interpretations and utilizes stability selection to identify meaningful and reproducible associations between brain-imaging modalities and behaviour. We apply this framework to exhibit specific brain-behaviour associations present in the transdiagnostic cohort Healthy Brain Network (HBN). The identified brain-behaviour associations establish connections between regional cortical features from structural magnetic resonance imaging and electronic clinical record forms assessing psychiatric symptoms. We show this framework is able to find relevant and stable associations
Mekkaoui, Mourad. "Outil de base en vue d'une approche de description d'informations en vision par ordinateur : Application aux tâches opératoires." Valenciennes, 1995. https://ged.uphf.fr/nuxeo/site/esupversions/0f570e9a-ec9a-4b93-a893-f529cfd94c26.
Dadi, Kamalaker. "Machine Learning on Population Imaging for Mental Health." Thesis, université Paris-Saclay, 2020. http://www.theses.fr/2020UPASG001.
Mental disorders display a vast heterogeneity across individuals. A fundamental challenge to studying their manifestations or risk factors is that the diagnosis of mental pathological conditions are seldom available in large public health cohorts. Here, we seek to develop brain signatures, biomarkers, of mental disorders. For this, we use ma-chine learning to predict mental-health outcomes through population imaging i. e. with brain imaging (Magnetic Resonance Imaging ( MRI )).Given behavioral or clinical assessments, population imaging can relate unique features of the brain variations to these non-brain self-reported measures based on questionnaires. These non-brain measurements carry a unique description of each individual’s psychological differences which can be linked to psychopathology using statistical methods. This PhD thesis investigates the potential of learning such imaging-based outcomes to analyze mental health. Using machine-learning methods, we conduct an evaluation, both a comprehensive and robust, of population measures to guide high-quality predictions of health outcomes. This thesis is organized into three main parts: first, we present an in-depth study of connectome biomarkers, second, we propose a meaningful data reduction which facilitates large-scale population imaging studies, and finally we introduce proxy measures for mental health. We first set up a thorough benchmark for imaging-connectomes to predict clinical phenotypes. With the rise in the high-quality brain images acquired without tasks, there is an increasing demand in evaluation of existing models for predictions. We performed systematic comparisons relating these images to clinical assessments across many cohorts to evaluate the robustness of population imaging methods for mental health. Our benchmarks emphasize the need for solid foundations in building brain networks across individuals. They outline clear methodological choices. Then, we contribute a new generation of brain functional atlases to facilitate high-quality predictions for mental health. Brain functional atlases are indeed the main bottleneck for prediction. These atlases are built by analyzing large-scale functional brain volumes using scalable statistical algorithm, to have better grounding for outcome prediction. After comparing them with state-of-the-art methods, we show their usefulness to mitigate large-scale data handling problems. The last main contribution is to investigate the potential surrogate measures for health outcomes. We consider large-scale model comparisons using brain measurements with behavioral assessments in an imaging epidemiological cohort, the United Kingdom ( UK ) Biobank. On this complex dataset, the challenge lies in finding the appropriate covariates and relating them to well-chosen outcomes. This is challenging, as there are very few available pathological outcomes. After careful model selection and evaluation, we identify proxy measures that display distinct links to socio-demographics and may correlate with non-pathological conditions like the condition of sleep, alcohol consumption and physical fitness activity. These can be indirectly useful for the epidemiological study of mental health
Sambra-Petre, Raluca-Diana. "2D/3D knowledge inference for intelligent access to enriched visual content." Electronic Thesis or Diss., Evry, Institut national des télécommunications, 2013. http://www.theses.fr/2013TELE0012.
This Ph.D. thesis tackles the issue of sill and video object categorization. The objective is to associate semantic labels to 2D objects present in natural images/videos. The principle of the proposed approach consists of exploiting categorized 3D model repositories in order to identify unknown 2D objects based on 2D/3D matching techniques. We propose here an object recognition framework, designed to work for real time applications. The similarity between classified 3D models and unknown 2D content is evaluated with the help of the 2D/3D description. A voting procedure is further employed in order to determine the most probable categories of the 2D object. A representative viewing angle selection strategy and a new contour based descriptor (so-called AH), are proposed. The experimental evaluation proved that, by employing the intelligent selection of views, the number of projections can be decreased significantly (up to 5 times) while obtaining similar performance. The results have also shown the superiority of AH with respect to other state of the art descriptors. An objective evaluation of the intra and inter class variability of the 3D model repositories involved in this work is also proposed, together with a comparative study of the retained indexing approaches . An interactive, scribble-based segmentation approach is also introduced. The proposed method is specifically designed to overcome compression artefacts such as those introduced by JPEG compression. We finally present an indexing/retrieval/classification Web platform, so-called Diana, which integrates the various methodologies employed in this thesis
Skjermo, Jo. "Case-based reasoning in medical image diagnosis." Thesis, Norwegian University of Science and Technology, Department of Computer and Information Science, 2001. http://urn.kb.se/resolve?urn=urn:nbn:no:ntnu:diva-262.
In the last several years, there has been an increased focus on connecting image processing and artificial intelligence. Especially in the field of medical image diagnostics the benefits for such integration is apparent. In this paper we present use of the Common Object Request Broker Architecture (CORBA), as the mean for connecting existing systems for image processing and artificial intelligence. To visualize this, we will use CORBA for connecting Dynamic Imager and JavaCreek. Dynamic Imager is an image processing software, that is especially suitable for setting up and test customized sequences of image processing operations. JavaCreek is an artificial intelligence software based on the Case-Based Reasoning (CBR) theory.
After connecting the two systems with CORBA, we proceed develop the specific image processing methods for data gathering, and a knowledge base for diagnosis in the artificial intelligence system. The image processing methods and the knowledge base are produced for one special knowledge domain, for visualizing how the proposed system can help in medical image diagnostics.
The task we use to visualize our approach, is detecting malignancy in breast tumors, from magnetic resonance (MR) images taken over time as contrast agents is injected. This is from a reasonably new method for deciding if a tumor is malignant or benign. All image processing methods and the knowledge base is produced to let the two systems cooperate to find and diagnose tumors.
The image processing methods, the knowledge model, and the selected software with CORBA connection, was the basis for our system implementation. The implementation was tested with data gathered during the development of the clinical method for determining if a tumor is malignant, from the MR images. In all 127 patient cases was available, where 77 has malignant tumors in the gathered images. The results was then compared with diagnosis methods based on manual detection, and on other image processing methods. Although the found results were promising, there was also found several areas for future work.
Kunda, Maithilee. "Visual problem solving in autism, psychometrics, and AI: the case of the Raven's Progressive Matrices intelligence test." Diss., Georgia Institute of Technology, 2013. http://hdl.handle.net/1853/47639.
Grand-Brochier, Manuel. "Descripteurs 2D et 2D+t de points d'intérêt pour des appariements robustes." Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2011. http://tel.archives-ouvertes.fr/tel-00697021.
Naffouti, Seif Eddine. "Reconnaissance et correspondance de formes 3D pour des systèmes intelligents de vision par ordinateur." Thesis, Bourgogne Franche-Comté, 2018. http://www.theses.fr/2018UBFCK033/document.
This thesis concerns recognition and matching of 3D shapes for intelligent computer vision systems. It describes two main contributions to this domain. The first contribution is an implementation of a new shape descriptor built on the basis of the spectral geometry of the Laplace-Beltrami operator; we propose an Advanced Global Point Signature (AGPS). This descriptor exploits the intrinsic structure of the object and organizes its information in an efficient way. In addition, AGPS is extremely compact since only a few eigenpairs were necessary to obtain an accurate shape description. The second contribution is an improvement of the wave kernel signature; we propose an optimized wave kernel signature (OWKS). The refinement is with a modified particle swarm optimization heuristic algorithm to better match a query to other shapes belonging to the same class in the database. The proposed approach significantly improves the discriminant capacity of the signature. To assess the performance of the proposed approach for nonrigid 3D shape retrieval, we compare the global descriptor of a query to the global descriptors of the rest of shapes in the dataset using a dissimilarity measure and find the closest shape. Experimental results on different standard 3D shape benchmarks demonstrate the effectiveness of the proposed matching and retrieval approaches in comparison with other state-of-the-art methods
Delcourt, Alexandre. "Amélioration des détecteurs CdZnTe pour l'imagerie gamma par apprentissage." Electronic Thesis or Diss., Université Grenoble Alpes, 2023. http://www.theses.fr/2023GRALM056.
Since a few years, the wide spread use of CZT-based detectors in gamma imaging drives their performance optimization to stay competitive at the industrial level. However, the presence of structural defect in the CZT crystal deteriorates the output signals quality and holds back the higher volume detectors development.The purpose of this thesis is the use of optimization and artificial intelligence algorithms using realistic simulations to override the impact of the defects and improve the localization performances of gamma interactions in the detector. We will develop a mathematical-based method in three steps as an alternative to common characterization and correction methods.First, we develop 3D CZT detector simulations enabling to implement defects with different natures to observe their impact on output signals. Then we build a simple neural network, which can be introduced in the electronics to localize the gamma interactions in the detector from simulation results. A second network based on a gradient computation method will allow determining the electric field and collection performance of a detector.The addition of these three steps will be used to learn through simulation the intern parameters of a determined detector such as the electric field. This simulation will serve to train the simple neural network and finally be used on experimental data to improve the localization performance of the detector.The development of this mathematical approach will help us having a better understanding of the intern structure of a CZT crystal being able to reproduce its behavior in simulation. In addition, the better performance of the detector might be sufficient to decrease the radiotracer dose for medical imaging or limit the exposition time of operators in a nuclear power plant