Geologically young gullies have been proposed to be evidence of recent liquid water on Mars. This dissertation details work I have done to address issues surrounding the Martian gullies and recent water on Mars. In order to determine the elevations at which gullies occur, I created a set of Interactive Data Language programs and Unix C-shell scripts to coregister Mars Orbiter Laser Altimeter topography with high resolution Mars images. My scripts represent the first public method that does this. Recently, the Mars Orbiter Camera detected changes in the form of new bright deposits in two gullies. The High Resolution Imaging Science Experiment (HiRISE) camera detected more gullies with bright deposits. I used my scripts to identify some of the best candidates for liquid water formation based on their shallow average slopes. A Digital Elevation Model (DEM) was produced using HiRISE stereo images of my selected candidates in Hale Crater. I model two gullies with bright slope deposits in Hale Crater and find that both water-rich and sediment-rich flows could reproduce the bright deposits’ locations and morphologies. Since liquid water is rarely stable on Mars today, I suggest that dry flows formed the bright deposits. The channel gradient where flows deposit, the apex slope, can tell us whether a flow was likely dry and non-fluidized (slopes ~21°) or fluidized (shallower slopes). I measured the apex slope of 75 gullies located in five HiRISE DEMs. I find that 72% of the gullies studied were likely emplaced by a fluidized flow. I also find that modified gullies are more likely to have a fluidized emplacement than relatively fresh gullies. My results suggest that there is no evidence requiring water-rich flows in gullies today. Understanding the concept of water on Mars is crucial to understanding NASA’s Mars Exploration Strategy, “Follow the Water.” I undertook a study investigating alternative conceptions about water on Mars held by middle school science teachers to understand what they hear when scientists say “water on Mars.” All study participants had alternative conceptions about water on Mars. I suggest focus topics for space science education programs.

BACKGROUND : Radiation oncology was these 20 last years revolutionized by the 3-dimensional conformal radiotherapy (3D-CRT) and its technical evolution, the intensity modulated radiotherapy (IMRT). Thanks to steep dose gradient dose distribution, these techniques allow to conform the prescribed dose to the Planning Target Volume (PTV) while significantly decreasing the dose delivered to the Organs at Risk (OAR). One critical step remains the accurate definition of the Gross Tumor Volume (GTV). If the GTV is underestimated, there is a risk of missing part of the target. If the GTV is overestimated, the risk is to overirradiate normal tissues. Today's gold standard for GTV definition is the Computed Tomography (CT) scanner. We though know that its poor soft tissues contrast is a factor of variability for target definition purpose. AIMS : It can be hypothesized that, for Head and Neck Squamous Cell Carcinomas located in the oropharynx or the laryngo-hypopharynx, the use of other anatomical (like Magnetic Resonance Imaging – MRI) or functional (like positron emission tomography with either 11C-methionine – MET-PET- or 18F-fluorol-deoxy-glucose – FDG-PET) imaging modalities could complement CT for GTV delineation, and have an impact on subsequent CTV and PTV delineation and dose distribution to the non target tissues outside the PTV. RESULTS : We could demonstrate that, providing an adequate and controlled methodology concerning image coregistration and tumor volume delineation on functional images, differences were observed for the delineation of primary tumor volume or GTV according to the modality used. Moreover, the trends were the same for both locations studied (oropharyngeal and laryngo-hypopharyngeal) : CT, MRI and MET-PET volumes were not significantly different in absolute volumes, but there was no total overlap, each imaging modality having the tendency to visualize different types and relatively specific pathways of tumor extension (e.g. : cartilages in MRI). What was very interesting was the significantly smaller FDG-PET volume which could have a real impact on radiation oncology practice by (1) allowing to reduce dose distribution and (2) providing fast and reproducible GTV delineation based on its functional characteristic. Furthermore, we could demonstrate on the subset of operated patients that these smaller FDG-PET volumes were not the fact of a volume underestimating delineation algorithm but well the reflection of true tumor extension. But one must keep in mind that because of spatial resolution limitations, there was still a significant overestimate of this true GTV. Also, none of the imaging modalities was able to visualize very small tumor extensions. This last fact put in the light the need for strict guidelines for CTV prediction based on GTV extension. This is what was done with the help of both anatomical and histo-pathological literature data. These guidelines were used to delineate CTVs on our images, allowing to perform comparative planning on primary tumor. It could be concluded that differences in GTV had not only an impact on CTV and subsequent PTV, but also on dose distribution, either on total irradiated volume or -perhaps more important- on mean dose to parotid glands. No significant effect could be observed on maximal dose to spinal cord. Compared to planning performed on macroscopy-based volumes, no significant difference could be found with what was done on PET-derived planning. CONCLUSION : This research paves the way for the use of FDG-PET for GTV delineation in planning the patients with oropharyngeal and laryngo-hypopharyngeal squamous cell carcinomas. / INTRODUCTION : La radiothérapie moderne a terriblement évolué ces 20 dernières années grâce au développement de la radiothérapie conformationnelle tridimensionnelle (3D-CRT) et de son évolution technique, la radiothérapie par modulation d'intensité (IMRT). Grâce à la création de gradients de dose très raides, ces techniques permettent de conformer au mieux la distribution de la dose au “Planning Target Volume” (PTV) tout en diminuant de manière significative la dose délivrée aux Organes à Risque (OAR). La précision de la définition du “Gross Tumor Volume” (GTV) ou volume tumoral macroscopique reste une étape cruciale dans le sens où une sous-estimation du volume augmente le risque de sous-doser la dose délivrée à la tumeur. Dans l'autre sens, la surestimation du volume tumoral conduit immanquablement à une surirradiation des tissus sains. La tomographie computée par scanner (CT) est l'imagerie de référence pour la définition du GTV. Cependant, le manque de constraste entre tissus mous – à fortiori entre la tumeur et les tissus environnants- constitue un facteur de variabilité reconnu quant à la précision de délimitation du GTV. BUTS : Pour les cancers de la sphère cervico-maxillo-faciale, en particulier pour les tumeurs épithéliales oropharyngées et laryngo-hypopharyngées, démontrer que l'usage complémentaire d'une autre imagerie anatomique comme la résonance magnétique (IRM) ou fonctionnelle comme la tomographie par émission de positrons utilisant soit la méthionine marquée au carbone 11 (MET-TEP), soit le fluoro-déoxy-glucose marqué au fluor 18 (FDG-TEP) peut améliorer la précision de la délimitation GTV. Dans ce cas, démontrer également que cela a un impact sur la délimitation des CTV et PTV sous-jacents et, in fine, sur la distribution de la dose aux tissus sains extérieurs au PTV. RESULTATS : Moyennant l'utilisation adéquate et contrôlée de méthodes de corégistration des images et de délimitation automatique des volumes en imagerie fonctionnelle, nous avons pu démontrer des différences en terme de GTV délimité selon les différentes modalités d'imagerie, avec une tendance identique que l'on se situe au niveau oropharyngé ou laryngo-hypopharyngé. Les GTV délimités sur CT, IRM et MET-TEP n'étaient pas significativement différents en valeurs absolues, mais chaque modalité avait tendance, au-delà d'une zone de congruence s'élevant en moyenne à 50% du volume total, à visualiser des extensions vers des zones anatomiques lui étant propre (ex. : les cartilages en IRM). Les volumes délimités en FDG-TEP étaient significativement plus petits que ceux délimités sur les autres modalités d'imagerie. Nous pûmes de plus démontrer sur un ensemble de patients opérés par laryngectomie totale que le FDG-TEP était aussi la plus précise des modalités d'imagerie. Cependant, par manque de résolution spatiale, aucune des modalités d'imagerie ne fut en mesure de couvrir totalement le GTV. Ce fait met en lumière le besoin de recommendations claires pour la prédiction du CTV sur base de l'extension du GTV. Ce travail fut réalisé sur base des données de la littérature anatomique (normale et pathologique). Ces recommendations furent utilisées pour délimiter les CTV sur les images CT, FDG-TEP et du spécimen chirurgical (les imageries IRM et MET-TEP ne furent pas analysées puisque n'apportant rien en regard du CT). Les PTV furent ensuite générés et une planification tridimensionnelle réalisée. Tant les CTV que les PTV délimités sur le FDG-TEP restaient significativement plus petits que leurs homologues délimités sur CT. Cette réduction permettait une réduction de la dose délivrée aux glandes parotides en particulier, aux tissus hors PTV de manière plus générale. CONCLUSION : Cette recherche ouvre la voie à l'utilisation du FDG-TEP pour la délimitation du GTV chez les patients atteints de tumeurs épithéliales des sphères oropharyngée et laryngo-hypopharyngée.

Coregistration of airborne laser scanning data and aerial images Abstract This thesis is dealing with the co-registration of aerial laser scanning and aerial images. Theoretical part with research of current methods puts emphasis on methods suitable for remote sensing datasets. Part of the thesis is about pre-processing data for co-registration and DSM production. Selection of co- registration methods for remote sensing is based on previous researches. Selected co-registration methods are applied on datasets from EuroSDR research project and ČÚZK dataset. Application is realised by programming codes and functions that were created for this purpose in Matlab. Possibilities of usage, advantages and disadvantages of methods are being mentioned in the next parts of the thesis with emphasis on time of the computation and final accuracy. The function programmed in Matlab allows comparison of co-registration methods and allows the user to decide which of the co- registration methods to use on input datasets. Discussion section describes the possibilities of method extensions and problematic parts across the whole co-registration process. Keywords: co-registration, laser, scanning, images, photogrammetry, remote sensing, coordinate, image matching.

A integração de diferentes modalidades de imagens médicas possibilita uma análise mais detalhada de seu conteúdo, visando-se um diagnóstico mais preciso da patologia presente. Este processo, conhecido como corregistro, busca o alinhamento das imagens através da transformação rígida (ou não rígida) das mesmas, por algoritmos matemáticos de distorção, translação, rotação e ajuste de escala. A amplitude de cada transformação é determinada por uma medida de similaridade das imagens. Quanto menor a similaridade, maior será a transformação aplicada. Neste sentido, a métrica de similaridade é uma peça chave do processo de corregistro. No presente trabalho, inicialmente são propostas novas definições para o cálculo dos erros de alinhamento nas transformações de translação, rotação e escala, com o objetivo de se avaliar o desempenho do corregistro. Em seguida, cinco experimentos são realizados. No primeiro, a Informação Mútua Generalizada é avaliada como medida de similaridade para corregistro em imagens fractais e cerebrais. Neste caso, os resultados sugerem a viabilidade do emprego desta métrica, pois em geral conduz a erros de alinhamento muito pequenos, mas sem vantagens aparentes em relação à formulação de Shannon. No segundo experimento, um estudo comparativo entre a Informação Mútua e as métricas clássicas (Coeficiente de Correlação, Média dos Quadrados, Diferença de Gradiente e Cardinalidade) é então realizado. Para as imagens binárias analisadas, as métricas com menores valores de erro de alinhamento para os corregistros de translação e rotação foram a Informação Mútua e a Diferença de Gradiente. Para o corregistro de escala, todas as métricas conduziram a erros de alinhamento próximos de zero. No terceiro experimento, o processo de alinhamento é investigado em termos do número de iterações do algoritmo de corregistro. Considerando-se ambas as variáveis erro de alinhamento e número de iterações, conclui-se que o uso da Informação Mútua Generalizada com q = 1.0 é adequado ao corregistro. No quarto experimento, a influência da dimensão fractal no corregistro de imagens fractais binárias foi estudada. Para algumas métricas, a tendência geral observada é a de uma diminuição do erro de alinhamento em resposta ao aumento da dimensão fractal. Finalmente, no quinto experimento, constatou-se a existência de correlação linear entre os erros de alinhamento de imagens em tons de cinza do córtex cerebral e de fractais do conjunto Julia. / The integration of different modalities of medical images provides a detailed analysis of its contents, aiming at a more accurate diagnosis of the pathology. This process, known as coregistration, seeks to align the images through rigid (or non-rigid) transformations, by mathematical algorithms of distortion, translation, rotation and scaling. The amplitude of each transformation is determined by a similarity measure of the images. The lower the similarity, the greater the transformation applied. In this sense, the similarity metric is the key for the coregistration process. In this work, new definitions are proposed for the calculation of alignment errors in the transformations of translation, rotation and scale, with the objective of evaluating the performance of coregistration. Then, five experiments are performed. In the first one, the Generalized Mutual Information is evaluated as a similarity measure for coregistration of brain images and fractals. In this case, the results suggest the feasibility of using this measure, since it leads to very small alignment errors, although no advantages in relation to Shannon formulation are evident. In the second experiment, a comparative study between Mutual Information and the classical metrics (Correlation Coefficient, Mean Squares, Gradient Difference and Cardinality) is performed. For the binary images analyzed, the metrics with lower alignment errors for translation and rotation are the Mutual Information and Gradient Difference. For scaling transformation, all the metrics lead to alignment errors close to zero. In the third experiment, the alignment process is investigated in terms of number of iterations of the coregistration algorithm. Considering both variables alignment error and number of iterations, it is concluded that the use of Generalized Mutual Information with q =1 is appropriate for coregistration. In the fourth experiment, it is studied the influence of fractal dimension in coregistration of binary fractal images. For some metrics, as a general trend, one observes the decay of the alignment error in response to the increase of the fractal dimension. Finally, in the fifth experiment, the results indicate the existence of a linear correlation between the alignment errors of grayscale images of the cerebral cortex and Julia set fractals.

<div>The Sentinel-1 (S1) mission is a part of the European Space Agency (ESA) Copernicus program. In 2014 and 2016, the mission launched the twin Synthetic Aperture Radar (SAR) satellites, Sentinel-1A (S1A) and Sentinel-1B (S1B). The S1 mission has started a new era for earth observations missions with its higher spatial resolution, shorter revisit days, more precise control of satellites orbits and the unprecedented free-to-public distribution policy. More importantly, S1 adopts a new wide swath mode, the TOPS mode as it default acquisition mode. The TOPS mode scans several different subswaths for gaining a larger coverage. Because the S1 mission is aimed at earth observation missions, for example, earthquakes, oods, ice sheets flow, etc., thus it is desired to have large monitoring areas. Although TOPS is still a relatively new idea, the high quality data and wide application scopes from S1 has earned tremendous attention in the SAR community.</div><div><br></div><div><div>The signal properties of wide swath mode such as TOPS are different from the more conventional stripmap mode, and it requires special techniques for successfully processing such data in the sense of interferometry. For the purpose of doing Interferometric SAR (InSAR), the coregistration step is of most critical because it requires a 1/1000 accuracy. In addition, processing wide swath mode requires special steps such as bursts stitching, deramping and reramping, and so on. Compared with stripmap, the processing techniques of wide swath mode are less developed. Much exploitations are still needed for how to design a generic and robust wide swath interferometric</div><div>processing chain.</div></div><div><br></div><div><div>Driven by the application needs of S1 wide swath interferometric processing, this research studies the key methodologies, explores and implements new processing chain, designs a generic wide swath processing </div><div>flow that would utilize the existing stripmap processing platform, as well as carries out preliminary applications. For studying key methods, this study carries out a quantitative analysis between two different coregistration methods, namely the cross-correlation approach and the geometrical</div><div>approach. The advantages and disadvantages for each method are given by the author, and it is proposed to choose the suitable method based on one's study area. For the implementation of the new processing chain, the author proposes a user-friendly stripmap-like processing </div><div>ow with all the wide swath related process done behind the scene. This approach allows people with basic knowledge in InSAR and very few knowledge in wide swath mode to be able to process and get interferometric products. For designing the generic process flow, the author applied TOPS's work flow to the other wide swath mode, ScanSAR mode and demonstrated the feasibility of processing two different wide swath mode with the same processing chain.</div><div>For preliminary applications, the author shows a large number of interferometric data throughout the research and presents a case study with multi temporal time series analysis using a stack of S1 dataset.</div></div><div><br></div><div><div>This research is application oriented, which means the study serves for real-world applications. Up to now, the processing chain and methodologies implemented in this</div><div>research has been shared by many research groups around the world and has seen a number of promising outcomes. The recognition from others is also an affrmation to the value of this research.</div></div>

A integração de diferentes modalidades de imagens médicas possibilita uma análise mais detalhada de seu conteúdo, visando-se um diagnóstico mais preciso da patologia presente. Este processo, conhecido como corregistro, busca o alinhamento das imagens através da transformação rígida (ou não rígida) das mesmas, por algoritmos matemáticos de distorção, translação, rotação e ajuste de escala. A amplitude de cada transformação é determinada por uma medida de similaridade das imagens. Quanto menor a similaridade, maior será a transformação aplicada. Neste sentido, a métrica de similaridade é uma peça chave do processo de corregistro. No presente trabalho, inicialmente são propostas novas definições para o cálculo dos erros de alinhamento nas transformações de translação, rotação e escala, com o objetivo de se avaliar o desempenho do corregistro. Em seguida, cinco experimentos são realizados. No primeiro, a Informação Mútua Generalizada é avaliada como medida de similaridade para corregistro em imagens fractais e cerebrais. Neste caso, os resultados sugerem a viabilidade do emprego desta métrica, pois em geral conduz a erros de alinhamento muito pequenos, mas sem vantagens aparentes em relação à formulação de Shannon. No segundo experimento, um estudo comparativo entre a Informação Mútua e as métricas clássicas (Coeficiente de Correlação, Média dos Quadrados, Diferença de Gradiente e Cardinalidade) é então realizado. Para as imagens binárias analisadas, as métricas com menores valores de erro de alinhamento para os corregistros de translação e rotação foram a Informação Mútua e a Diferença de Gradiente. Para o corregistro de escala, todas as métricas conduziram a erros de alinhamento próximos de zero. No terceiro experimento, o processo de alinhamento é investigado em termos do número de iterações do algoritmo de corregistro. Considerando-se ambas as variáveis erro de alinhamento e número de iterações, conclui-se que o uso da Informação Mútua Generalizada com q = 1.0 é adequado ao corregistro. No quarto experimento, a influência da dimensão fractal no corregistro de imagens fractais binárias foi estudada. Para algumas métricas, a tendência geral observada é a de uma diminuição do erro de alinhamento em resposta ao aumento da dimensão fractal. Finalmente, no quinto experimento, constatou-se a existência de correlação linear entre os erros de alinhamento de imagens em tons de cinza do córtex cerebral e de fractais do conjunto Julia. / The integration of different modalities of medical images provides a detailed analysis of its contents, aiming at a more accurate diagnosis of the pathology. This process, known as coregistration, seeks to align the images through rigid (or non-rigid) transformations, by mathematical algorithms of distortion, translation, rotation and scaling. The amplitude of each transformation is determined by a similarity measure of the images. The lower the similarity, the greater the transformation applied. In this sense, the similarity metric is the key for the coregistration process. In this work, new definitions are proposed for the calculation of alignment errors in the transformations of translation, rotation and scale, with the objective of evaluating the performance of coregistration. Then, five experiments are performed. In the first one, the Generalized Mutual Information is evaluated as a similarity measure for coregistration of brain images and fractals. In this case, the results suggest the feasibility of using this measure, since it leads to very small alignment errors, although no advantages in relation to Shannon formulation are evident. In the second experiment, a comparative study between Mutual Information and the classical metrics (Correlation Coefficient, Mean Squares, Gradient Difference and Cardinality) is performed. For the binary images analyzed, the metrics with lower alignment errors for translation and rotation are the Mutual Information and Gradient Difference. For scaling transformation, all the metrics lead to alignment errors close to zero. In the third experiment, the alignment process is investigated in terms of number of iterations of the coregistration algorithm. Considering both variables alignment error and number of iterations, it is concluded that the use of Generalized Mutual Information with q =1 is appropriate for coregistration. In the fourth experiment, it is studied the influence of fractal dimension in coregistration of binary fractal images. For some metrics, as a general trend, one observes the decay of the alignment error in response to the increase of the fractal dimension. Finally, in the fifth experiment, the results indicate the existence of a linear correlation between the alignment errors of grayscale images of the cerebral cortex and Julia set fractals.