No description available.

No description available.

No description available.

Målet med uppsatsen var att se hur man producerar fram en reklamkampanj, framför allt hur man gör det för första gången samt hur retoriken kan bidra till ett lyckat koncept. Vilka förutsättningar krävs? Arbetar man på ett annorlunda sätt? Vilka medel använder man för att få en framgångsrik reklam? Detta har undersökts genom att intervjua personer som varit med och producerat den prisbelönta och världskända reklamkampanjen från IKEA  ”Drömkök åt alla” ur ett retoriskt perspektiv. Resultatet visar på en resa rätt i tiden  i hur man med ny teknik får fram något helt unikt. / The purpose of this essay was to see how to produce a commercial campaign, and most of allhow to do it for the first time and how rhetoric’s can contribute to a successful concept. Whatare the conditions? Does it demand a different work approach? Which means do you use tocome up with a successful advertising? This has been investigated by interviewing the peoplewho produced the award winning and world famous commercial-campaign from IKEA"Dream kitchen for everyone" from a rhetoric point of view. The results display a journeymade at the right period of time, and how you with new technology, can come up withsomething unique.

Les méthodes de recalage d’images, qui ont pour but l’alignement de deux ou plusieurs images dans un même système de coordonnées, sont parmi les algorithmes les plus anciens et les plus utilisés en vision par ordinateur. Les méthodes de recalage servent à établir des correspondances entre des images (prises à des moments différents, par différents senseurs ou avec différentes perspectives), lesquelles ne sont pas évidentes pour l’œil humain. Un type particulier d’algorithme de recalage, connu comme « les méthodes de recalage déformables à l’aide de modèles graphiques » est devenu de plus en plus populaire ces dernières années, grâce à sa robustesse, sa scalabilité, son efficacité et sa simplicité théorique. La gamme des problèmes auxquels ce type d’algorithme peut être adapté est particulièrement vaste. Dans ce travail de thèse, nous proposons plusieurs extensions à la théorie de recalage déformable à l’aide de modèles graphiques, en explorant de nouvelles applications et en développant des contributions méthodologiques originales.Notre première contribution est une extension du cadre du recalage à l’aide de graphes, en abordant le problème très complexe du recalage d’une tranche avec un volume. Le recalage d’une tranche avec un volume est le recalage 2D dans un volume 3D, comme par exemple le mapping d’une tranche tomographique dans un système de coordonnées 3D d’un volume en particulier. Nos avons proposé une formulation scalable, modulaire et flexible pour accommoder des termes d'ordre élevé et de rang bas, qui peut sélectionner le plan et estimer la déformation dans le plan de manière simultanée par une seule approche d'optimisation. Le cadre proposé est instancié en différentes variantes, basés sur différentes topologies du graph, définitions de l'espace des étiquettes et constructions de l'énergie. Le potentiel de notre méthode a été démontré sur des données réelles ainsi que des données simulées dans le cadre d’une résonance magnétique d’ultrason (où le cadre d’installation et les stratégies d’optimisation ont été considérés).Les deux autres contributions inclues dans ce travail de thèse, sont liées au problème de l’intégration de l’information sémantique dans la procédure de recalage (indépendamment de la dimensionnalité des images). Actuellement, la plupart des méthodes comprennent une seule fonction métrique pour expliquer la similarité entre l’image source et l’image cible. Nous soutenons que l'intégration des informations sémantiques pour guider la procédure de recalage pourra encore améliorer la précision des résultats, en particulier en présence d'étiquettes sémantiques faisant du recalage un problème spécifique adapté à chaque domaine.Nous considérons un premier scénario en proposant un classificateur pour inférer des cartes de probabilité pour les différentes structures anatomiques dans les images d'entrée. Notre méthode vise à recaler et segmenter un ensemble d'images d'entrée simultanément, en intégrant cette information dans la formulation de l'énergie. L'idée principale est d'utiliser ces cartes estimées des étiquettes sémantiques (fournie par un classificateur arbitraire) comme un substitut pour les données non-étiquettées, et les combiner avec le recalage déformable pour améliorer l'alignement ainsi que la segmentation.Notre dernière contribution vise également à intégrer l'information sémantique pour la procédure de recalage, mais dans un scénario différent. Dans ce cas, au lieu de supposer que nous avons des classificateurs arbitraires pré-entraînés à notre disposition, nous considérons un ensemble d’annotations précis (vérité terrain) pour une variété de structures anatomiques. Nous présentons une contribution méthodologique qui vise à l'apprentissage des critères correspondants au contexte spécifique comme une agrégation des mesures de similarité standard à partir des données annotées, en utilisant une adaptation de l’algorithme « Latent Structured Support Vector Machine ». / Image registration methods, which aim at aligning two or more images into one coordinate system, are among the oldest and most widely used algorithms in computer vision. Registration methods serve to establish correspondence relationships among images (captured at different times, from different sensors or from different viewpoints) which are not obvious for the human eye. A particular type of registration algorithm, known as graph-based deformable registration methods, has become popular during the last decade given its robustness, scalability, efficiency and theoretical simplicity. The range of problems to which it can be adapted is particularly broad. In this thesis, we propose several extensions to the graph-based deformable registration theory, by exploring new application scenarios and developing novel methodological contributions.Our first contribution is an extension of the graph-based deformable registration framework, dealing with the challenging slice-to-volume registration problem. Slice-to-volume registration aims at registering a 2D image within a 3D volume, i.e. we seek a mapping function which optimally maps a tomographic slice to the 3D coordinate space of a given volume. We introduce a scalable, modular and flexible formulation accommodating low-rank and high order terms, which simultaneously selects the plane and estimates the in-plane deformation through a single shot optimization approach. The proposed framework is instantiated into different variants based on different graph topology, label space definition and energy construction. Simulated and real-data in the context of ultrasound and magnetic resonance registration (where both framework instantiations as well as different optimization strategies are considered) demonstrate the potentials of our method.The other two contributions included in this thesis are related to how semantic information can be encompassed within the registration process (independently of the dimensionality of the images). Currently, most of the methods rely on a single metric function explaining the similarity between the source and target images. We argue that incorporating semantic information to guide the registration process will further improve the accuracy of the results, particularly in the presence of semantic labels making the registration a domain specific problem.We consider a first scenario where we are given a classifier inferring probability maps for different anatomical structures in the input images. Our method seeks to simultaneously register and segment a set of input images, incorporating this information within the energy formulation. The main idea is to use these estimated maps of semantic labels (provided by an arbitrary classifier) as a surrogate for unlabeled data, and combine them with population deformable registration to improve both alignment and segmentation.Our last contribution also aims at incorporating semantic information to the registration process, but in a different scenario. In this case, instead of supposing that we have pre-trained arbitrary classifiers at our disposal, we are given a set of accurate ground truth annotations for a variety of anatomical structures. We present a methodological contribution that aims at learning context specific matching criteria as an aggregation of standard similarity measures from the aforementioned annotated data, using an adapted version of the latent structured support vector machine (LSSVM) framework.

Multiple Sclerosis (MS) is an inflammatory disease which causes areas of demyelination in the Central Nervous System (CNS) and affects only humans. Current therapies for MS are focused on anti-inflammatory treatment, which reduce the occurrence and clinical relapses of the disease. However, progressive disability of the disease is related to axonal degeneration. After demyelination, remyelination occurs, which helps repair the demyelinated lesions and protects axons from degeneration. However, this endogenous remyelination is inefficient, and currently there are no therapies available to enhance remyelination. The aim of this thesis was to first characterize a fast and reliable model to study CNS remyelination in vitro, and second to investigate the role of semaphorin 3a (Sema3A) and semaphorin 3f (Sema3F) signaling in CNS remyelination. Various in vivo models have been developed to investigate the pathology of multiple sclerosis, and can be used to test remyelination therapies. However, in vivo models are expensive, animal- and time- consuming. Until now, there has been no well-characterized and robust in vitro model for remyelination study. In this thesis, an ex vivo slice culture system with mouse brain and spinal cord was developed, and characterized by immunofluorescent microscopy and transmission electron microscopy, for CNS remyelination study. Automated (re)myelinating quantification by image pro plus software was developed and validated to provide a fast and reliable way for testing factors that change remyelination efficiency. Two such factors are Sema3A and 3F, which were initially identified as axon guidance cues during development. Sema3A (repulsive) and 3F (attractive) were proved to play a role in oligodendrocyte precursor cell (OPC) migration during development, and hypothesized to be important in remyelination. In this thesis, I investigated the effects and mechanisms for this by adding recombinant SEMA3A or SEMA3F or by knockdown their obligatory receptors Neuropilin (Nrp) 1 and 2, using lentivirus induced miRNAi. Slice culture and primary OPC culture were used to determine the effect on OPC survival, migration, proliferation, differentiation and myelination.

International Telemetering Conference Proceedings / October 17-20, 1988 / Riviera Hotel, Las Vegas, Nevada / The growing complexity of missle, aircraft, and space vehicle systems, along with the advent of fly-by-wire and ultra-high performance unstable airframe technology has created an exploding demand for real time processing power. Recent VLSI developements have allowed addressing these needs in the design of a multi-processor subsystem supplying 10 MIPS and 5 MFLOPS per processor. To provide up to 70 MIPS a Digital Signal Processing subsystem may be configured with up to 7 Processors. Multiple subsystems may be employed in a data processing system to give the user virtually unlimited processing power. Within the DSP module, communication between cards is over a high speed, arbitrated Private Data bus. This prevents the saturation of the system bus with intermediate results, and allows a multiple processor configuration to make full use of each processor. Design goals for a single processor included executing number system conversions, data compression algorithms and 1st order polynomials in under 2 microseconds, and 5th order polynomials in under 4 microseconds. The processor design meets or exceeds all of these goals. Recently upgraded VLSI is available, and makes possible a performance enhancement to 11 MIPS and 9 MFLOPS per processor with reduced power consumption. Design tradeoffs and example applications are presented.

In this paper we explore the slice-ribbon conjecture for some families of pretzel knots. Donaldson's diagonalization theorem provides a powerful obstruction to sliceness via the union of the double branched cover W of B⁴ over a slicing disk and a plumbing manifold P([capital gamma]). Donaldson's theorem classifies all slice 4-strand pretzel knots up to mutation. The correction term is another 3-manifold invariant defined by Ozsváth and Szabó. For a slice knot K the number of vanishing correction terms of Y[subscript K] is at least the square root of the order of H₁(Y[subscript K];Z). Donaldson's theorem and the correction term argument together give a strong condition for 5-strand pretzel knots to be slice. However, neither Donaldson's theorem nor the correction terms can distinguish 4-strand and 5-strand slice pretzel knots from their mutants. A version of the twisted Alexander polynomial proposed by Paul Kirk and Charles Livingston provides a feasible way to distinguish those 5-strand slice pretzel knots and their mutants; however the twisted Alexander polynomial fails on 4-strand slice pretzel knots. / text

Image-guided surgeries employ advanced imaging and computing technologies to assist the surgeon when direct visualization is inadequate or unavailable. As modern surgeries continue to move toward minimally invasive procedures, tracked ultrasound (US), an emerging technology that uniquely combines US imaging and position tracking, has been increasingly used for intraoperative guidance in surgical interventions. The intrinsic accuracy of a tracked US system is primarily determined by a unique procedure called ``probe calibration", where a spatial registration between the coordinate systems of the transducer (provided by a tracking device affixed to the probe) and the US image plane must be established prior to imaging. Inaccurate system calibration causes misalignments between the US image and the surgical end-effectors, which may directly contribute to treatment failure. The probe calibration quality is further reduced by the "elevation beamwidth" or "slice thickness", a unique feature of the ultrasound beam pattern that gives rise to localization errors and imaging uncertainties. In this thesis, we aim to provide an automated, pure-computation-based, intraoperative calibration solution that also incorporates the slice thickness to improve the calibration accuracy, precision and reliability. The following contributions have been made during the course of this research. First, we have designed and developed an automated, freehand US calibration system with instant feedback on its calibration accuracy. The system was able to consistently achieve submillimeter accuracy with real-time performance. Furthermore, we have developed a novel beamwidth-weighted calibration framework (USB-FW) that incorporates US slice thickness to improve the estimation of calibration parameters. The new framework provides an effective means of quality control for calibration results. Extensive phantom validation demonstrated that USB-FW introduces statistically significant reduction (p = 0.001) in the calibration errors and produces calibration outcomes that are less variable than a conventional, non-beamwidth-weighted calibration. Finally, we were the first to introduce an automated, intraoperative Transrectal Ultrasound (TRUS) calibration technology for needle guidance in prostate brachytherapy. Our tests with multiple commercial TRUS scanners and brachytherapy stepper systems demonstrated that the proposed method is practical in use and can achieve high calibration accuracy, precision and robustness. / Thesis (Ph.D, Computing) -- Queen's University, 2012-10-22 16:18:55.439

Thesis (B.S.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1982. / MICROFICHE COPY AVAILABLE IN ARCHIVES AND ENGINEERING / by Joseph Jay Kesselman Jr. / B.S.