Il y a plus de soixante-dix ans que la résonance magnétique nucléaire (RMN) a été découverte, mais elle est toujours prospère et vivante, couvrant un large éventail d'applications dans les sciences, technologies et industries. Une application omniprésente de la résonance magnétique nucléaire est une technique d'imagerie appelée imagerie par résonance magnétique (IRM), qui a trouvé beaucoup d'applications en médecine, sciences, et technologie. Les techniques de transformation de Fourier dites par ''encodage dans l’espace k'' sont des méthodes d'IRM basées sur l'acquisition d'un signal de résonance magnétique en fonction d’un paramètre "k" qui sera ensuite transformé en une image par transformation de Fourier. Aujourd'hui, les techniques de Fourier sont les plus importantes en IRM, mais il existe des alternatives parmi lesquelles ''l'encodage spatial'', qui est le sujet principal de cette thèse. Dans l’encodage spatial (également connu sous le terme d’encodage temporel ou encodage spatiotemporel), l'acquisition du signal s'effectue de telle manière que l'intensité du signal ressemble à l'objet. Par conséquent, dans l'encodage spatial, la transformation de Fourier n'est pas nécessaire pour la reconstruction de l'image.Il a été montré que les techniques d'imagerie hybride à balayage unique, qui utilisent l'encodage k traditionnel dans une direction et l'encodage spatiotemporel dans l'autre, sont supérieures aux méthodes traditionnelles qui utilisent l'encodage k dans les deux directions, notamment pour supprimer les effets de variations de fréquence (causées par des champs magnétique inhomogènes, ou par la présence de plusieurs déplacements chimiques, ou toute autre source de variations de fréquence), et conduisent à des images beaucoup moins déformées que les méthodes d'imagerie traditionnelles. Dans cette thèse, l'idée de l'imagerie par résonance magnétique par encodage spatial sera discutée. La formation de l'image et les propriétés des images résultant de différentes séquences d'encodage spatial seront brièvement étudiées.Les effets de la diffusion sur une séquence hybride établie appelée "acquisition rapide par excitation séquentielle et refocalisation" (RASER) sont étudiés. On montrera que dans les séquences d'encodage spatial, l'atténuation du signal due à la diffusion n'est souvent pas uniforme sur l’ensemble de l'objet, provoquant un contraste trompeur dans l'image. Afin d'éliminer ce faux contraste, une séquence d'impulsion comprenant deux impulsions balayées en fréquence (DC-RASER) est proposée. Les résultats expérimentaux sont conformes à nos prévisions théoriques sur les effets de la diffusion dans ces séquences. Ils confirment que l'atténuation du signal due à la diffusion est uniforme sur l’ensemble de l’objet.Afin de développer les applications de l'encodage spatial à balayage unique, nous montrons comment on peut améliorer le contraste dans la séquence originale RASER. En changeant le déroulement de la séquence d'impulsions, nous avons réalisé une variante de RASER appelée RASER avec écho décalé (ES-RASER), qui fournit un niveau de contraste réglable.Enfin, nous montrons comment on peut améliorer quelques aspects des séquences à encodage temporel disponibles. En réarrangeant les gradients positifs et négatifs, nous montrons comment on peut réduire la vitesse de la commutation des gradients. Ceci est important, car une commutation rapide des gradients n'est pas toujours techniquement possible et peut en plus stimuler involontairement le système nerveux du patient. En utilisant un gradient supplémentaire, nous avons pu modifier l'ordre de détection dans la séquence originale d’encodage temporel. Cela conduit à un temps d'écho identique pour tous les échos, et à une atténuation uniforme du signal due à la relaxation. Finalement, nous montrons comment on peut répartir l’acquisition des séquences d'encodage temporel de façon entrelacée, afin de réduire l'atténuation du signal due à la diffusion. / Although Nuclear Magnetic Resonance (NMR) has been discovered more than seventy years ago, it is still thriving and alive, covering a broad spectrum of applications in science, technology and industry. One of the most ubiquitous applications of Nuclear Magnetic Resonance is an imaging technique dubbed Magnetic Resonance Imaging (MRI), which has found many applications in science, technology, and particularly in medicine. Fourier or k-encoding techniques are MRI methods based on acquiring a magnetic resonance signal as a function of the parameter “k”, a subsequent Fourier transform then will convert the signal to an image. Although nowadays Fourier techniques are prominent in MRI, there are other alternatives, among which spatial encoding, the main subject of this dissertation, should be mentioned. In spatial encoding (also known as time- or spatiotemporal-encoding), signal acquisition is performed in such a way that the signal intensity itself resembles the object. Consequently, in spatial encoding there is no need for a Fourier transform for image reconstruction.Single-scan hybrid imaging techniques that use traditional k-encoding in one direction, and spatial (time-)encoding in the other have been shown to be superior to traditional full k-encoding methods (that use k-encoding in both directions) in suppressing the effects of frequency variations (caused by inhomogeneous magnetic fields, the presence of more than one chemical shift, or any other frequency variation) and lead to images that are much less distorted than traditional single-scan imaging methods. In this dissertation the main idea behind spatial encoding magnetic resonance imaging will be introduced. Image formation and image properties in different spatial encoding sequences will also be briefly investigated.Then, the effects of diffusion on an established hybrid sequence called “Rapid Acquisition by Sequential Excitation and Refocusing, RASER” are investigated. It will be shown that in spatial encoding sequences, the attenuation of the signal due to diffusion is often not uniform across the entire object, leading to misleading contrast in the image. In order to eliminate this misleading contrast, a double-chirp RASER (DC-RASER) pulse sequence is proposed in this work. The experimental results are in accordance with our theoretical investigations about the effects of diffusion in these sequences. They also confirm that the signal attenuation due to diffusion is uniform, as expected theoretically for DC-RASER.In order to develop applications of single-scan spatial encoding MRI we show how one can enhance the contrast in the original RASER sequence. By changing the timing of the pulse sequence, we achieved a variant of RASER called Echo-Shifted RASER (ES-RASER), which provides a tunable contrast level.Finally, we show how one can improve a few aspects of the available time-encoding sequences. By rearranging positive and negative gradients we show how one can reduce the switching rate of the gradients. This is important because fast gradient switching is not always technically feasible; in addition, it may unwittingly stimulate the patient’s nervous system. By using an additional gradient we can change the detection order in the original time-encoding sequence. This leads to an identical echo-time for all echoes, and hence a uniform signal attenuation due to relaxation. Furthermore, we show how one can implement time-encoding sequences in an interleaved fashion in order to reduce signal attenuation due to diffusion.

The purpose of this study is to analyze two works within the fantasy and the fantastic genre, Eragon by Christopher Paolini and Border by John Ajvide Lindqvist, and how ideas of race are expressed in either work. This study will also discuss how racial elements in fantasy can be utilized in education, both to problematize those ideas as well as how they can be used as a resource to gain some new perspectives from a didactic perspective. This study aims to answer the following question formulations: How does either work express racial ideas? What similarities and differences can be found in their expressions of race? How can these racial ideas be problematized and used as a resource from a didactic perspective? Critique of ideology is the overarching method, while the analytic methods used are close reading and comparative analysis. The theory behind the study is how race are constructed from a norm, humans, and how other races in a fantastical setting differ from this norm by being directly compared to the norm. This also dictates what constitutes as a human or “other”, as well as were leakage can be found in the norms. The conclusion of this study is that both of these works being analyzed shows different aspects of how fantasy and the fantastical can be used in education when it comes to the racial ideas, one upholds the racial ideas of the past and the other uses those same ideas for subversion. In education, works within the fantasy and the fantastical genre can be utilized in two different ways. The first is to problematize and create awareness of the racial ideas in fantasy literature, and the second is to use fantasy works that subverts racial ideas in order to deconstruct and gain a new understanding of race and racism in our own world.

Magnetic resonance imaging (MRI) is a critical diagnostic tool in medical practice, enabling non-invasive visualization of anatomy and physiological processes. Nonetheless, MRI has inherent spatial resolution limitations, which may limit its diagnostic capabilities. Recently, a new technology employing Radio-frequency Amplification by Stimulated emission of Radiation (RASER) has emerged to improve MRI resolution. Similar to a laser, RASER-MRI signals spontaneously emerge without the need for a radio frequency pulse(RF), which additionally enhances the safety of the process. However, RASER-MRI images frequently exhibit a significant presence of image artifacts due to the nonlinear behavior between image slices. This master’s thesis aims to determine whether image artifacts can be eliminated using deep artificial neural networks. The neural networks were trained on purely synthetic data, due to the complexity of real RASER experiments. The implementation was split into three phases. The first phase focused on the reconstruction of 1D RASER profiles. The test done during this phase showed that the reconstruction was preferably made with a Convolutional Neural Network (CNN). The CNN does not require knowledge of the total population inversion, and the ideal input was the most volatile RASER spectrum. The second phase was dedicated to reconstructing simulated RASER-MRI images. This phase started with the creation of a random RASER-MRI image generator which was used to generate the training and testing data. The reconstruction was successful and was further enhanced with an image-to-image Unet. The entire deep learning pipeline did not suffice for real data, which sparked the third phase. The third phase focused on simulating more realistic RASER data. The new data improved the result, however, the reconstruction did not suffice. Further research needs to be done into ways to make the simulation more realistic to improve the reconstruction of the real RASER-MRI image. However, this project concludes that simulated RASER-spectra can be reconstructed using deep learning.