No description available.

L'imagerie ultrasonore est une modalité d'acquisition privilégiée en imagerie médicale en raison de son innocuité, sa simplicité d'utilisation et son coût modéré d'utilisation. Néanmoins, la résolution limitée et le faible contraste limitent son utilisation dans certaines d'applications. C'est dans ce contexte que différentes techniques de post-traitement visant à améliorer la qualité de telles images sont proposées dans ce manuscrit. Dans un premier temps, nous proposons d'aborder le problème conjoint de la déconvolution et de la segmentation d'images ultrasonores en exploitant l'interaction entre ces deux problèmes. Le problème, énoncé dans un cadre bayésien, est résolu à l'aide d'un algorithme MCMC en raison de la complexité de la loi a posteriori des paramètres d'intérêt. Dans un second temps, nous proposons une nouvelle méthode rapide de super-résolution fondée sur la résolution analytique d'un problème de minimisation l2-l2. Il convient de remarquer que les deux approches proposées peuvent être appliquées aussi bien à des images ultrasonores qu'à des images naturelles ou constantes par morceaux. Enfin, nous proposons une méthode de déconvolution aveugle basée sur un modèle paramétrique de la réponse impulsionelle de l'instrument ou du noyau de flou. / In the field of medical image analysis, ultrasound is a core imaging modality employed due to its real time and easy-to-use nature, its non-ionizing and low cost characteristics. Ultrasound imaging is used in numerous clinical applications, such as fetus monitoring, diagnosis of cardiac diseases, flow estimation, etc. Classical applications in ultrasound imaging involve tissue characterization, tissue motion estimation or image quality enhancement (contrast, resolution, signal to noise ratio). However, one of the major problems with ultrasound images, is the presence of noise, having the form of a granular pattern, called speckle. The speckle noise in ultrasound images leads to the relative poor image qualities compared with other medical image modalities, which limits the applications of medical ultrasound imaging. In order to better understand and analyze ultrasound images, several device-based techniques have been developed during last 20 years. The object of this PhD thesis is to propose new image processing methods allowing us to improve ultrasound image quality using postprocessing techniques. First, we propose a Bayesian method for joint deconvolution and segmentation of ultrasound images based on their tight relationship. The problem is formulated as an inverse problem that is solved within a Bayesian framework. Due to the intractability of the posterior distribution associated with the proposed Bayesian model, we investigate a Markov chain Monte Carlo (MCMC) technique which generates samples distributed according to the posterior and use these samples to build estimators of the ultrasound image. In a second step, we propose a fast single image super-resolution framework using a new analytical solution to the l2-l2 problems (i.e., $\ell_2$-norm regularized quadratic problems), which is applicable for both medical ultrasound images and piecewise/ natural images. In a third step, blind deconvolution of ultrasound images is studied by considering the following two strategies: i) A Gaussian prior for the PSF is proposed in a Bayesian framework. ii) An alternating optimization method is explored for blind deconvolution of ultrasound.

Digital holography is a technique for the recording of holograms via CCD/CMOS devices and enables their subsequent numerical reconstruction within computers, thus avoiding the photographic processes that are used in optical holography. This thesis investigates the various techniques which have been developed for digital holography. It develops and successfully demonstrates a number of refinements and additions in order to enhance the performance of the method and extend its applicability. The thesis contributes to both the experimental and numerical analysis aspects of digital holography. Regarding experimental work: the thesis includes a comprehensive review and critique of the experimental arrangements used by other workers and actually implements and investigates a number of these in order to compare performance. Enhancements to these existing methods are proposed, and new methods developed, aimed at addressing some of the perceived short-comings of the method. Regarding the experimental aspects, the thesis specifically develops:• Super-resolution methods, introduced in order to restore the spatial frequencies that are lost or degraded during the hologram recording process, a problem which is caused by the limited resolution of CCD/CMOS devices.• Arrangements for combating problems in digital holography such as: dominance of the zero order term, the twin image problem and excessive speckle noise.• Fibre-based systems linked to tunable lasers, including a comprehensive analysis of the effects of: signal attenuation, noise and laser instability within such systems.• Two-source arrangements for contouring, including investigating the limitations on achievable accuracy with such systems. Regarding the numerical processing, the thesis focuses on three main areas. Firstly, the numerical calculation of the Fresnel-Kirchhoff integral, which is of vital importance in performing the numerical reconstruction of digital holograms. The Fresnel approximation and the convolution approach are the two most common methods used to perform numerical reconstruction. The results produced by these two methods for both simulated holograms and real holograms, created using our experimental systems, are presented and discussed. Secondly, the problems of the zero order term, twin image and speckle noise are tackled from a numerical processing point of view, complementing the experimental attack on these problems. A digital filtering method is proposed for use with reflective macroscopic objects, in order to suppress both the zero-order term and the twin image. Thirdly, for the two-source contouring technique, the following issues have been discussed and thoroughly analysed: the effects of the linear factor, the use of noise reduction filters, different phase unwrapping algorithms, the application of the super-resolution method, and errors in the illumination angle. Practical 3D measurement of a real object, of known geometry, is used as a benchmark for the accuracy improvements achievable via the use of these digital signal processing techniques within the numerical reconstruction stage. The thesis closes by seeking to draw practical conclusions from both the experimental and numerical aspects of the investigation, which it is hoped will be of value to those aiming to use digital holography as a metrology tool.

abstract: Image resolution limits the extent to which zooming enhances clarity, restricts the size digital photographs can be printed at, and, in the context of medical images, can prevent a diagnosis. Interpolation is the supplementing of known data with estimated values based on a function or model involving some or all of the known samples. The selection of the contributing data points and the specifics of how they are used to define the interpolated values influences how effectively the interpolation algorithm is able to estimate the underlying, continuous signal. The main contributions of this dissertation are three fold: 1) Reframing edge-directed interpolation of a single image as an intensity-based registration problem. 2) Providing an analytical framework for intensity-based registration using control grid constraints. 3) Quantitative assessment of the new, single-image enlargement algorithm based on analytical intensity-based registration. In addition to single image resizing, the new methods and analytical approaches were extended to address a wide range of applications including volumetric (multi-slice) image interpolation, video deinterlacing, motion detection, and atmospheric distortion correction. Overall, the new approaches generate results that more accurately reflect the underlying signals than less computationally demanding approaches and with lower processing requirements and fewer restrictions than methods with comparable accuracy. / Dissertation/Thesis / Ph.D. Bioengineering 2013

Standard optical microscopes cannot resolve images below 200 nm within the visible wavelengths due to optical diffraction limit. This Thesis reports an investigation into super-resolution imaging beyond the optical diffraction limit by microsphere optical nano-scopy (MONS) and submerged microsphere optical nano-scopy (SMON). The effect of microsphere size, material and the liquid type as well as light illumination conditions and focal plane positions on imaging resolution and magnification have been studied for imaging both biological (viruses and cells) and non-biological (Blu-ray disk patterns and nano-pores of anodised aluminium oxide) samples. In particular, sub-surface imaging of nano-structures (data-recorded Blu-ray) that cannot even be seen by a scanning electron microscope (SEM) has been demonstrated using the SMON technique. Adenoviruses of 75 nm in size have been observed with white light optical microscopy for the first time. High refractive index microsphere materials such as BaTiO3 (refractive index n = 1.9) and TiO2-BaO-ZnO (refractive index n = 2.2) were investigated for the first time for the imaging. The super-resolution imaging of sub-diffraction-limited objects is strongly influenced by the relationship between the far-field propagating wave and the near-field evanescent waves. The diffraction limit free evanescent waves are the key to achieving super-resolution imaging. This work shows that the MONS and SMON techniques can generate super-resolution through converting evanescent waves into propagating wave. The optical interactions with the microspheres were simulated using special software (DSIMie) and finite different in time domain numerical analysis software (CST Microwave Studio). The optical field structures are observed in the near-field of a microsphere. The photonic nanojets waist and the distance between single dielectric microsphere and maximum intensity position were calculated. The theoretical modelling was calculated for comparisons with experimental measurements in order to develop and discover super-resolution potential.

Différentes études précédentes ont démontré que IL-1β et PKA peuvent réduire la transmission synaptique inhibitrice dans la LAMINA II de la moelle épinière, en contribuent de cette manière au développement de douleur chronique de tipe inflammatoire. Au niveau des sites post-synaptiques, les changements dans la transmission synaptique (par exemple suivant le relâchement de IL-1β ou après l’activation de PKA), reflètent donc des changements dans les propriétés et/ou dans le nombre des molécules présentes au niveau de la synapse. Au cours de mon doctorat, j’ai pu profiter des techniques basés sur l’imagerie des molécules uniques afin d’étudier les effets de PKA et IL-1β sur la dynamiques et le nombre absolu de GlyR dans les synapses de la moelle épinière. Mes résultats ont montré que PKA et Il-1β peuvent déplacer les GlyR des sites inhibitoires post-synaptiques ciblent différentes sous-unités α du récepteur de la glycine. Comme les sous-unités GlyRα ne se lient pas directement à la géphyrine, ces effets sont vraisemblablement le résultat d’un changement de conformation du GlyR dépendant de la sous-unité. Pendant mon projet, j’ai utilisé la technique de microscopie de super-résolution PALM pour développer une méthode pour déterminer la stœchiométrie des GlyR et le nombre absolu de récepteurs et des molécules d’échafaudage au niveau des synapse de la moelle épinière. Mes résultats décrivent que les GlyR se composent de 3 sous-unités α et de 2 sous-unités β, et proposent qu’une synapse de la moelle épinière contient en moyenne 80 GlyR et 250 molécules de géphyrine. Ces résultats sont essentiels pour mettre en relation l’ampleur des mécanismes de régulation et de plasticité agissant sur la transmission synaptique, avec les changements en nombre de molécules présentes dans les synapses de la moelle épinière. Sur la base de mes découvertes on pourra maintenant étudier les mécanismes structuraux impliqués dans la régulation de la fonction et la dynamique des GlyR dépendantes des sous-unités α que j’ai démontré. / IL-1β and PKA impair glycine receptor-mediated synaptic transmission in the lamina II of the spinal cord, contributing to the development of inflammatory types of chronic pain. At post-synaptic sites, the strength of synaptic transmission depends on the biophysical properties and on the absolute number of receptors expressed. Consequently, changes in synaptic transmission (i.e. following the release of IL-1β or the activation of PKA), reflect changes in the properties and/or number of molecules present at the synapse. During my PhD I have taken advantage of single-molecule based imaging techniques to study the effects of IL-1β and PKA on the dynamics and absolute numbers of GlyRs at spinal cord synapses.My results show for the first time that both Il-1β and PKA displace GlyRs from inhibitory post-synaptic sites, targeting different α-subunit of GlyRs. Specifically, IL-1β reduces GlyR α-containing receptors at spinal cord synapse, whereas PKA affects GlyR α3L subunit. Given that the GlyR α subunits do not bind to the gephyrin scaffold, these effects likely reflect an α-subunit dependent change in GlyR conformation that decreases the affinity of the GlyR subunits for gephryrin. Glycine receptors are composed of α- and β- subunits that assemble into heteropentameric complexes with an unclear stoichiometry. Using super resolution PALM microscopy I have developed a single-molecule counting approach to determine the stoichiometry of GlyRs and the absolute number of receptor and scaffold molecules at spinal cord synapses. According to my results GlyRs is composed by 3 α and 2 β-subunits, and an average spinal cord synapse contains around 80 GlyRs and 250 scaffold molecules. These data are fundamental to relate the magnitude of regulatory and plasticity mechanisms acting on glycinergic transmission, with quantitative changes in molecule numbers at spinal cord synapses. My research has shown how absolute quantitative approaches can help achieve a more detailed insight into the organization of complex molecular assemblies and their dynamic regulation.

This work applies sparse representations and nonlinear image processing to two inverse imaging problems. The first problem involves image restoration, where the aim is to reconstruct an unknown high-quality image from a low-quality observed image. Sparse representations of images have drawn a considerable amount of interest in recent years. The assumption that natural signals, such as images, admit a sparse decomposition over a redundant dictionary leads to efficient algorithms for handling such sources of data. The standard sparse representation, however, does not consider the intrinsic geometric structure present in the data, thereby leading to sub-optimal results. Using the concept that a signal is block sparse in a given basis —i.e., the non-zero elements occur in clusters of varying sizes — we present a novel and efficient algorithm for learning a sparse representation of natural images, called graph regularized block sparse dictionary (GRBSD) learning. We apply the proposed method towards two image restoration applications: 1) single-Image super-resolution, where we propose a local regression model that uses learned dictionaries from the GRBSD algorithm for super-resolving a low-resolution image without any external training images, and 2) image inpainting, where we use GRBSD algorithm to learn a multiscale dictionary to generate visually plausible pixels to fill missing regions in an image. Experimental results validate the performance of the GRBSD learning algorithm for single-image super-resolution and image inpainting applications. The second problem addressed in this work involves image enhancement for detection and segmentation of objects in images. We exploit the concept that even though data from various imaging modalities have high dimensionality, the data is sufficiently well described using low-dimensional geometrical structures. To facilitate the extraction of objects having such structure, we have developed general structure enhancement methods that can be used to detect and segment various curvilinear structures in images across different applications. We use the proposed method to detect and segment objects of different size and shape in three applications: 1) segmentation of lamina cribrosa microstructure in the eye from second-harmonic generation microscopy images, 2) detection and segmentation of primary cilia in confocal microscopy images, and 3) detection and segmentation of vehicles in wide-area aerial imagery. Quantitative and qualitative results show that the proposed methods provide improved detection and segmentation accuracy and computational efficiency compared to other recent algorithms.

Multi-frame image super-resolution focuses on reconstructing a high-resolution image from a set of low-resolution images with high similarity. Since super-resolution is an ill-posted problem, regularization techniques are widely used to constrain the minimization function. Combining image prior knowledge with fidelity model, Bayesian-based methods can effectively solve this ill-posed problem, which makes this kind of methods more popular than other methods. Our proposed model is based on maximum a posteriori probability (MAP) estimation. In this thesis, we propose a novel initialization method based on median operator to initialize our estimated high-resolution image. For the fidelity term in our proposed algorithm, the half-quadratic estimation is used to choose error norm adaptively instead of using fixed L1 or L2 norm. Furthermore, for our regularization term, we propose a novel regularization method based on Difference Curvature (DC) and Bilateral Total Variation (BTV) to suppress mixed noises and preserve image edges simultaneously. In our experimental results, synthetic data and real data are both tested to demonstrate the superiority of our proposed method in terms of clearer texture and less noise over other state-of-the-art methods.

Video super-resolution (VSR) aims to recover a realistic high-resolution (HR) frame from its corresponding center low-resolution (LR) frame and several neighbouring supporting frames. The neighbouring supporting LR frames can provide extra information to help recover the HR frame. However, these frames are not aligned with the center frame due to the motion of objects. Recently, many video super-resolution methods based on deep learning have been proposed with the rapid development of neural networks. Most of these methods utilize motion estimation and compensation models as preprocessing to handle spatio-temporal alignment problem. Therefore, the accuracy of these motion estimation models are critical for predicting the high-resolution frames. Inaccurate results of motion compensation models will lead to artifacts and blurs, which also will damage the recovery of high-resolution frames. We propose an effective wide activated separate 3 dimensional (3D) Convolution Neural Network (CNN) for video super-resolution to overcome the drawback of utilizing motion compensation models. Separate 3D convolution factorizes the 3D convolution into convolutions in the spatial and temporal domain, which have benefit for the optimization of spatial and temporal convolution components. Therefore, our method can capture temporal and spatial information of input frames simultaneously without additional motion evaluation and compensation model. Moreover, the experimental results demonstrated the effectiveness of the proposed wide activated separate 3D CNN.

Nanoscopic spatial reorganization of selectin ligands, CD44 and PSGL-1, during the initial step of hematopoietic stem/progenitor cell (HSPC) homing, tethering and rolling of migrating cells over E-selectins, has been recently reported. However, the exact spatial distribution of these ligands and their spatial reorganization during the cell rolling on E-selectins are still an open question. The spatiotemporal characterization at the nanoscale level requires high resolution imaging methods. In this study, I quantitatively characterize nanoscopic spatiotemporal behavior of the selectin ligands on the migrating cells to understanding the molecular mechanism of the cell rolling at the nanoscale level by means of a microfluidics-based 3D super-resolution fluorescence microscopy technique. The obtained results suggest that PSGL-1 on the cell shows significant change in the axial distribution on the cell during the cell rolling on E-selectin whereas the spatial distribution of CD44 along the axial direction is not affected significantly by the cell rolling. These findings indicate that each selectin ligand has a distinct contribution to the initial step of the HSPC homing because of their distinct spatial localizations on the cells that regulate at least partly the accessibility of these ligands to the surface E-selectin.