Brandt, Martin B.C. 2011. Imaging the African Superplume – Upper mantle, Tomography and Moment tensor. Ph.D. thesis, Faculty of Science, University of the Witwatersrand, Johannesburg, South Africa. The African Superplume, African Superswell and East African Rift System are amongst the most prominent geophysical features on Earth, but the structure, evolution and interaction between these features is controversial. In my thesis I conducted a range of investigations in an effort to better understand these issues. The thesis presents the investigations into the structure and expressions of these features. These include: (I) A study of the upper mantle shear velocity structure beneath southern Africa to investigate the source of the buoyancy that has powered the Superswell; (II) Statistical hypothesis testing of middle-mantle shear velocity tomographic models to evaluate evidence for links between the Superplume and low velocity features in/near the transition zone; and (III) Computation of three new regional moment tensors for South Africa to assess crustal stress in the Kalahari craton, and its link with mantle structure and dynamics. Waveform data were obtained for the study on the upper mantle shear velocity structure and the moment tensor inversions from the Southern African Seismic Experiment Kaapvaal craton array. For the statistical hypothesis testing on global tomography images, new travel-time data from both global and AfricaArray stations were added to Grand’s global shear velocity data set. The principal findings of this study are summarized below. I. The upper mantle shear velocity structure beneath the Kalahari craton is similar to that of other shields, except for slightly slower velocities from 110–220 km depth. The difference may be due to higher temperatures or a decrease in magnesium number (Mg#). If the slower velocities in the deep lithosphere are due solely to a temperature anomaly, then slightly less than half of the unusually high elevation of the Kalahari craton can be explained by shallow buoyancy from a depleted hot lithosphere. Decreasing the Mg# of the lower lithosphere would increase density and counteract higher temperatures. If an excess temperature of 90 K over a 110 km depth range and a corresponding decrease in Mg# of -2 between the Kalahari and the other cratons are assumed, this would match the seismic velocity difference but would result in essentially no buoyancy difference. We conclude that the high elevation of the Kalahari craton can only be partially supported by shallow mantle buoyancy and must have a deeper source. We determined a thickness of 250±30 km for the mantle transition zone below eastern southern Africa, which is similar to the global average, but the corresponding velocity gradient is less steep than in standard global models (PREM and IASP91). Velocity jumps of 0.16±0.1 km/s (eastern) and 0.21±0.1 km/s (central) across the 410 km discontinuity were found. Our results indicate a thermal or chemical anomaly in the mantle transition zone, but this cannot be quantified due to uncertainty. II. Statistical hypothesis testing on our global tomography images indicated that the African Superplume rises from the core-mantle boundary to at least 1150 km depth, and the upper mantle slow-velocity anomaly extends from the base of the lithosphere to below the mantle transition zone. The model that links the African Superplume with the slow-velocity anomaly in the upper mantle under eastern Africa has an equal probability to an alternative hypothesis with a thin slow-velocity “obstruction zone” at 850 to 1000 km depth. III. Finally, we calculated three regional moment tensors for South Africa and made progress towards resolving the discrepancy between the local and moment magnitudes we observe for the region. Moment tensors/focal mechanisms in southern Africa change from normal faulting (extension) in the northeast near the East African Rift to strike-slip faulting in the southwest. This confirms previous studies stating that not only eastern Africa, but also southern Africa is being actively uplifted by lithospheric modification at its base and/or the African Superplume.

A research report submitted to the Faculty of Health Sciences in fulfilment of the requirements for the degree of Master of Medicine, in Internal Medicine at the University of Witwatersrand, Johannesburg. September 2017 / Differentiating ischaemic left ventricular dysfunction (ILVD) from non-ischaemic left ventricular dysfunction (NILVD) is crucial since appropriately selected patients may benefit from coronary revascularisation. The aim of this study was to evaluate the diagnostic utility of myocardial perfusion imaging (MPI) in patients presenting with left ventricular dysfunction using coronary angiography (CA) as the gold standard. Methods This single centre retrospective study was conducted in 52 patients with heart failure with a reduced ejection fraction (EF< 40%) who had both MPI as well as CA at CHBAH between January 2005 and December 2012. ILVD was diagnosed when the distribution and severity of coronary disease on CA was sufficient to account for the degree of left ventricular dysfunction. Results From a total of 52 patients, 33 (63%) had ILVD and 19 (37%) had NILVD. As compared to patients with NILVD, those with ILVD were more likely to be Indian and White (p=0.0014), have more coronary risk factors (5(2) vs 3(2), p < 0.0001) and more commonly have q waves on the ECG (0% vs 55%, p < 0.0001). MPI had a sensitivity of 100% (95% CI 66-100%) and specificity of 52.63% (95% CI 30.18 - 75.08) for the diagnosis of ILVD. The presence of fixed perfusion defects on MPI was the best predictor of ILVD. Conclusion MPI has high sensitivity but low specificity for the diagnosis of ILVD. This makes it a useful screening test for the exclusion of coronary artery disease in patients presenting with heart failure. / MT2018

Abstract Medical x-ray imaging is used to visualise patients’ anatomical structures and in some cases their physiology. X-rays are ionizing radiation, thus their use needs to be optimised, as stochastic effects are assumed to increase linearly with the exposure dose. Imaging protocols need to be optimised to a radiation dose level that follows the as low as reasonably achievable principle without compromising the diagnostic value of the image. Different methods can be used to help in the optimisation process, such as simulations, radiation dose and image quality assessments with dosimeters and phantoms and utilising the latest technology in the most efficient way. The purpose of this doctoral thesis was to investigate the applicability of metal-oxide-semiconductor-field-effect-transistor (MOSFET) dosimeters for dose determinations in conventional x-ray and computed tomography (CT) examinations. Additionally, dose optimising methods were investigated in dental panoramic imaging using radiophotoluminescence (RPL) dosimeters. Anthropomorphic phantoms were used in every study to simulate patients, as their structures enable dosimeters to be positioned at locations that correspond to different organs. The MOSFET’s properties for dose determinations were evaluated against the reference dosimeter in a conventional x-ray set-up. Comparisons of absorbed and effective doses in thorax x-ray imaging were made between RPLs, MOSEFTs and Monte Carlo simulations. The effect of the organ-based tube current modulation and bismuth shields were compared against the reference imaging method in a chest CT with one scanner model. Absorbed doses and quantitative image quality were evaluated using each method. Possible dose reduction from segmented dental panoramic tomography (sDPT) imaging was compared against full DPT. Dose measurements were done using RPL dosimeters in pediatric and adult set-up using phantoms. MOSFETs are accurate enough to be used in conventional x-ray and CT, but they require a careful calibration before use as their reproducibility is limited with low doses. Bismuth shields provided the best dose reduction, but with a negative impact on quantitative image quality, especially when metal artefact removal software was used. The final study showed that the use of sDPT programmes and pediatric protocols enable a notably dose reduction compared to the full DPT adult protocol. / Tiivistelmä Lääketieteellisessä kuvantamisessa käytetään röntgensäteilyä potilaan anatomian ja joissain tapauksissa fysiologian visualisointiin. Röntgensäteily on ionisoivaa ja stokastisten vaikutusten kasvaessa oletettavasti lineaarisesti säteilyn funktiona, tulee säteilyn olla kokonaisvaltaisesti optimoitua. Kuvauksissa käytetyn röntgensäteilyn käytön tulee noudattaa ALARA-periaatetta, minkä vuoksi kuvauksessa tulee käyttää niin vähän säteilyä kuin vain mahdollista, diagnostiikan vaarantumatta. Optimoinnin apuna voidaan käyttää esim. simulointeja, annos- ja kuvanlaatumäärityksiä dosimetreilla ja fantomeilla, tai laitevalmistajien tuomia uusia teknologioita. Tämän väitöskirjan tarkoituksena oli tutkia metallioksidi-puolijohdekanavatransistorien (MOSFET) soveltuvuutta natiiviröntgentutkimuksissa ja tietokonetomografiassa (TT). Lisäksi työssä tutkittiin hammaskuvauksissa käytettyjä annossäästömenetelmiä radiofotoluminesenssidosimetreilla (RPL). Potilasvasteena työssä käytettiin antropomorfisia fantomeita, minkä ansiosta säteilyannoksia voidaan mitata eri elimiä vastaavilta kohdilta. MOSFET annosmittarin ominaisuuksia arvioitiin natiiviröntgenasetelmassa referenssimittariin nähden. Absorboituneiden ja efektiivisten annosten eroa MOSFET:tien, RPL:ien ja simulointien kesken tutkittiin keuhkoröntgentutkimuksessa. Pintakudoksia säästävän putkivirranmodulointimenetelmän ja vismuttisuojien vaikuttavuutta verrattiin TT:ssä referenssimetelmää vasten. Vaikuttavuutta arvioitiin absorboituneiden annosten ja kvantitatiivisen kuvanlaadun avulla. Segmentoidun hammaspanoraamakuvauksen (sDPT) annossäästömahdollisuuksia verrattiin tavalliseen panoraamakuvaukseen. Annosmääritykset tehtiin käyttäen RPL dosimetreja lapsi- ja aikuisfantomeissa. MOSFET dosimetreja voidaan käyttää annosmäärityksiin natiiviröntgenkuvauksissa ja TT:ssä, mutta niiden kalibrointi ja toistettavuus matalilla annoksilla aiheuttaa kuitenkin rajoituksia niiden käytölle. Vismuttisuojat tuottivat parhaan annossäästön, huonontaen kuitenkin kuvanlaatua. Kuvanlaadun huonontuminen oli erityisen huomattavaa, kun metallista aiheutuvien kuvavirheiden poistamiseen suunniteltua ohjelmaa käytettiin. Viimeinen tutkimus osoitti, että sDPT ohjelmat ja lapsille suunnatut protokollat mahdollistavat huomattavan annossäästön verrattuna aikuisten kokopanoraamaan.

L’objectif de ce travail est de positionner l’imagerie terahertz comme un nouvel outil dédié à la métrologie et au contrôle non destructif. Ses propriétés étonnantes à la fois de forte pénétration dans les matériaux diélectriques et de longueur d’onde de dimension millimétrique voire submillimétrique en font un outil très enthousiasmant pour des applications comme le contrôle non destructif (CND).En premier lieu, nous avons présenté deux techniques d’imagerie : une consistant à déplacer l’objet, point par point, dans un faisceau terahertz focalisé pour reconstruire, pixel par pixel, une image en deux dimensions. L’autre technique repose sur l’utilisation d’un objectif et d’un capteur matriciel à l’état de l’art fonctionnant à température ambiante. Ainsi nous avons imagé des objets d’intérêts, sans déplacement mécanique. Bien que cette dernière technique soit beaucoup plus rapide que l’imagerie point par point, la qualité des images obtenues n’est pas comparable et ceci à cause de deux verrous. Ainsi, une partie du travail a consisté en l’étude des stratégies d’éclairage pour limiter les effets optiques interférentiels induits par la cohérence de la source. De plus, en déployant des simulations de toute la chaine optique avec le logiciel Zemax, de nombreuses images de qualité permettent d’envisager des applications en imagerie 2D (métrologie, CND) et en imagerie 3D. Ainsi, deux techniques complémentaires de reconstruction tomographique sont expérimentées sur des images obtenues en temps réel ; une technique inspirée de l’imagerie par rayon X et une technique utilisée en microscopie optique 3D qui exploite la profondeur de champ réduite de l’objectif. Enfin, plusieurs études spectroscopiques et métrologiques sont menées afin d’évaluer le comportement des matériaux et la dimension des objets à partir de leurs images THz en 2D ou avec leur reconstruction tomographique 3D. Les travaux effectués constituent les briques de base au déploiement de l’imagerie terahertz vers le domaine applicatif grâce à une qualité d’image incomparable et des acquisitions en temps réels. / The aim of this work is to position terahertz imaging as a new tool dedicated to metrology and non-destructive testing. Its astonishing properties of both high penetration in dielectric materials and wavelength millimeter or even submillimetric make it a very exciting tool for applications such as non-destructive testing (NDT).First, we presented two imaging techniques: one to move the object, point by point, in a focused terahertz beam to reconstruct, pixel by pixel, a two-dimensional image. The other technique is based on the use of a lens and a matrix sensor in the state of the art operating at room temperature. Thus we have imaged objects of interest, without mechanical displacement. Although this last technique is much faster than point-by-point imaging, the quality of the images obtained is not comparable and this because of two drawbacks. Thus, part of the work consists in the study of lighting strategies to limit the interferential optical effects induced by the coherence of the source. In addition, by deploying simulations of the entire optical chain with Zemax software, numerous quality images make it possible to consider applications in 2D (metrology, NDT) and 3D imaging. Thus, two complementary tomographic reconstruction techniques are tested on images obtained in real time; a technique inspired by X-ray imaging and a technique used in 3D optical microscopy that exploits the reduced depth of field of the lens. Finally, several spectroscopic and metrological studies are carried out in order to evaluate the behavior of the materials and the dimension of the objects starting from their THz images in 2D or with their 3D tomographic reconstruction. The work done is the foundation for deploying terahertz imagery to the application domain with unmatched image quality and real-time acquisition.

Des mesures de résistivité électrique et des campagnes d’observations géologiques ont été menées pour cartographier le contact sédiment/roche basale ainsi que les structures d’impact associées au cratère d’impact Bosumtwi datant de 1,07 millions d’années. Le cratère de 10,5 km de diamètre s’est formé dans des roches métamorphiques du Précambrien (2,1 à 2,2 milliard d’années), d’origine sédimentaire et volcanique. Il est actuellement rempli par le lac Bosumtwi de 8,5 km de diamètre. Ce cratère est la source des tektites et microtektites dispersées jusqu’en Côte d’Ivoire et au large des côtes ouest africaine. La campagne de mesures éophysiques consiste en 16 tomographies de résistivité électrique effectuées radialement des rives du lac vers les bords du cratère d’impact. Chaque profil utilise un système d’acquisition multi – électrodes avec une distance minimum entre les électrodes de 5 m. Les données ont été corrigées des effets topographiques et inversées en utilisant le programme commercial d’inversion Res2DInv, avec la norme L1 considérée plus robuste. La zone comprise entre les rives du lac et le bord externe du cratère de divise en trois formations géologiques principales. Les régions de faibles résistivités (< 64 .m) représentent les sédiments de lac. Les zones de resistivités moyennes (entre 128 et 200 .m) sont interprétées comme des brèches liées à l’impact, en dikes, allochtones ou parautochtones. Les régions de hautes résistivités (> 128 .m) représentent les roches métamorphiques sous-jacentes, d’origine volcanique ou sédimentaire. Les profils de résistivités permettent de retrouver la géométrie et l’extension latérale de ces trois types de roches. Une correspondance directe entre la lithologie observée en surface et les structures mises en évidence par les mesures de résistivité électrique dans le sous-sol est observée à Dwamam au Sud-Est dulac. À Dwamam, les sédiments sont environ à 200 m de la rive du lac et s’étendent environ sur 400 m vers les les bords du cratère, à la différence des autres zones où ont pu être cartographiés les sédiments. La topographie du contact sédiment/roche basale présente une direction particulière NE-SO avec un pendage variant entre 16 degrés au NE et 36 degrés au SO. Une majorité de fractures ont été remarquées dans le SO, remplies par des clasts et des brèches d’impacts. Les failles sont surtout présentent à l’Ouest du lac. En moyenne elles présentent un pendage variant entre 60o à l’Est et 80 degrés à l’Ouest du cratère. Des analyses statistiques ont été effectuées sur les directions et les pendages des failles selon les loi de von Mises et Fisher. Elles démontrent que les failles s’alignent préférentiellement le long de deux directions principales. L’analyse cumulée de la surface de contact sédiment/roche, des failles et de la localisation du champ de dispersion des tektites indique que le bolide d’environ 0,8 à 1 km de diamètre responsable du cratère est arrivé du NE. Les résultats démontrent que les panneaux de résistivité électrique fournissent des informations utiles pour l’étude des cratères d’impact. / Electrical resistivity tomography (ERT) and geological field surveys have been used to map the sediment/bedrock contact and impact related structures of the 1.07 Myr old Bosumtwi impact crater. The 10.5 km complex crater excavated in 2.1–2.2 Gyr Precambrian metasedimentary and metavolcanic rocks is filled by the 8.5 km Lake Bosumtwi. It is the source crater of the tektites and microtektites of the Ivory Coast strewn field. Electrical resistivity survey was carried out sixteen (16) profiles running from the shore of the lake towards the rim of the crater. The multi-electrode gradient array method with minimum electrode separation of 5 m was used. The data were corrected for topography and inverted using the L1–norm (robust inversion) techniqueof the Res2DInv software. The area extending from the lake shore towards the crater rim contains essentially three formations. The low resistivity regions (< 64 .m) represent the lake sediments. The moderately high resistivity regions with values between 128 and 200 .m were interpreted as impact related breccias (dikes, allochthonous or parautochthonous) depending on their geometries. Lastly, the model clearly differentiate the resistive basement metamorphic rocks (> 128 .m) fromthe lake sediments and the breccias due to their geometry and lateral extent. Also observed was a direct correspondence between the lithology on the surface and the subsurface resistivity structures at Dwamam in the southeast section of the lake. At Dwamam, the sediments were about 200 m away from the shore and stretch about 400 m towards the crater rim unlike in other areas where the sediments were mapped from the shore. The gradient of the sediments/bedrock contact showsa symmetry in the NE–SW direction and dips between the lowest of 16 degrees in the NE to the highest 36 degrees in the SW. Majority of the fractures marked were in the southwest and were filled with clasts or impact breccia matrix. The faults were mostly delineated in the west. Averagely, the dips of the faults are about 60 degrees and 80 degrees for the east and west sections of the crater respectively. The dips of the faults were statistically treated using the von Mises and Fisher statistics, it was found that the faults have a preferred direction and it is possible to determine at least two different orientations. Theanalysis of the results of the sediments/bedrock surface and the faults combined with the location of the tektite strewn field indicate that the about 0.8–1 km bolide that created the complex crater came from the NE. The findings have shown that the ERT is efficient and a useful tool in impact cratering science research.

O tratamento cirúrgico das afecções que acometem a coluna cervical alta colocam em risco estruturas neurovasculares, como a artéria vertebral e a medula espinhal, durante a colocação de parafusos transarticulares em C1-C2 e parafusos de massa lateral em C1. O presente estudo visa determinar os parâmetros anatômicos do atlas (C1) em uma amostra da população brasileira, compará-los aos resultados já apresentados na literatura para outras populações e, assim, identificar e alterar as indicações dos implantes utilizados no tratamento das afecções da coluna cervical alta. Foi realizado um estudo observacional retrospectivo de um banco de dados prospectivo, incluindo 100 pacientes atendidos em um hospital terciário, entre janeiro de 2012 a dezembro de 2013. Foram utilizados cortes axiais e sagitais de tomografias computadorizadas. Os parâmetros estudados foram, no corte axial: o ponto de entrada do parafuso (DTPPE), o ângulo de segurança (ADSPA), o tamanho ideal do parafuso (TPA) e o tamanho da massa lateral entre os forames vertebral e transverso (DCVFT); no corte sagital foi medido o tamanho ideal do parafuso (TPS), o ângulo de segurança (ADSPS) e a espessura do arco posterior (EAP). Todos os parâmetros foram divididos de acordo com a idade, sexo e o lado esquerdo e direito. Resultados: O ponto de entrada dado pelo DTPPE foi 21,86 ± 1,5 mm a esquerda e 22,7 ± 1,44 mm a direita no sexo masculino; e 20 ± 1,4 mm a esquerda e 20,24 ± 1,34 mm a direita no sexo feminino. A zona de segurança (ADSPA) foi 23,68 ± 6,12º a esquerda e 24,0 ± 5,82º a direita no sexo masculino e 18,09 ± 5,46º a esquerda e 18,57 ± 5,34º a direita no sexo feminino. A espessura do arco posterior (EAP) encontrada no sexo masculino foi 8,95 ±1,75 mm a esquerda e 8,92 ± 2,22 mm a direita, no sexo feminino encontramos uma espessura menor de 7,21 ± 1,53 mm a esquerda e 7,41 ± 1,58 a direita. Por meio de uma técnica original e reprodutível de avaliação de exames de tomografia computadorizada, os parâmetros anatômicos do Atlas encontrados na amostra da população estudada em nosso estudo foram semelhantes aos apresentados previamente na literatura. No entanto, foram observadas diferenças entre os sexos. Assim, recomendamos a utilização do exame de tomografia computadorizada como método de avaliação pré-operatória quando a fixação da massa lateral pelo arco posterior do Atlas for empregada. / The surgical treatment of conditions affecting the high cervical spine puts at risk neurovascular structures, such as the vertebral artery and the spinal cord, during the placement of transarticular screws in C1-C2 and lateral mass screws in C1. The present study aims to determine the anatomical parameters of the atlas (C1) in a sample of the brazilian population, to compare them with the results already presented in the literature for other populations and thus to identify and change the indications of the implants used in the treatment of spinal affections cervical high. A retrospective observational study of a prospective database including 100 patients treated at a tertiary hospital between January 2012 and December 2013 was performed. Axial and sagittal sections of computed tomography were used. The parameters studied were, in the axial section: the screw entry point (DTPPE), the safety angle (ADSPA), the ideal screw size (TPA) and the lateral mass size between the vertebral and transverse foramina (DCVFT) ; in the sagittal cut, the ideal screw size (TPS), the safety angle (ADSPS) and the posterior arch thickness (EAP) were measured. All parameters were divided according to age, sex and left and right side. The entry point given by the DTPPE was 21.86 ± 1.5 mm on the left and 22.7 ± 1.44 mm on the right in the male; and 20 ± 1.4 mm on the left and 20.24 ± 1.34 mm on the right in the female. The safety zone (ADSPA) was 23.68 ± 6.12 ° on the left and 24.0 ± 5.82 ° on the right in the male sex and 18.09 ± 5.46 ° on the left and 18.57 ± 5.34 ° on the right in females. The posterior arch thickness (EAP) found in males was 8.95 ± 1.75 mm on the left and 8.92 ± 2.22 mm on the right, in females we found a thickness less than 7.21 ± 1.53 mm on the left and 7.41 ± 1.58 on the right. Using an original and reproducible technique for the evaluation of CT scans, the Atlas anatomical parameters found in the sample of the population studied in our study were similar to those previously reported in the literature. However, differences between the sexes were observed. Thus, we recommend the use of computed tomography (CT) examination as a preoperative evaluation method when lateral mass fixation by the posterior arch of the Atlas is used.

As lesões vasculares estão entre as anormalidades congênitas e neonatais mais comuns. Cerca de 60% estão localizadas na região da cabeça e pescoço. A identificação e a classificação adequadas de cada tipo de lesão, tem um impacto direto no diagnóstico e, no tratamento destes pacientes. A técnica de Tomografia por Coerência Óptica (Optical coherence tomography: OCT) obtém imagens em alta resolução, geradas em tempo real que permitem visualizar de forma não invasiva, estruturas celulares e extracelulares, com uma resolução espacial da ordem de 10&mu;m, que é 20 vezes maior que a de um ultrassom, aproximando-se do nível histológico convencional, razão pela qual é considerada biópsia óptica. Com o objetivo de fornecer imagens que possibilitassem avaliar as diferenças na arquitetura dos vasos em diferentes tipos de lesões localizadas, no tecido cutâneo e na mucosa, em regiões da cabeça e pescoço, foram tomadas imagens de lesões de 27 pacientes usando um sistema de OCT com Doppler (1325nm). Os resultados deste estudo clínico comprovaram a viabilidade da OCT permitindo distinguir entre a pele normal e a pele com lesão. Na má-formação capilar, os vasos sanguíneos tiveram diâmetro médio de 114\\&mu;m &plusmn; 92&mu;m e apresentaram uma profundidade média de 304 &mu;m &plusmn; 99 &mu;m. Os hemangiomas apresentaram vasos sanguíneos com diâmetro médio de 39&mu;m &plusmn; 19 &mu;m e profundidade média de 298 &mu;m &plusmn; 133 &mu;m. Nas lesões diagnosticadas pelo clínico como Hemangioma da mucosa labial, medimos diâmetro médio dos vasos de 45&mu;m &plusmn; 16 &mu;m com profundidade média de 520,5&mu;m &plusmn; 189,5&mu;m. Na má-formação linfática de lábio, os vasos sanguíneos apresentaram diâmetro médio de 157,5&mu;m &plusmn; 90&mu;m e, uma profundidade média 695&mu;m &plusmn; 136&mu;m. Na má-formação vascular de lábio, os vasos sanguíneos apresentaram diâmetro médio de 128&mu;m&plusmn;63&mu;m e, uma profundidade média 689 &mu;m &plusmn; 217&mu;m. Com estas informações obtidas das imagens de OCT indicando fluxo, foi possível avaliar diferenças peculiares da lesão de cada paciente, promovendo o diagnóstico diferencial das lesões vasculares da superfície tecidual da pele e mucosa oral. / Vascular lesions are among the most common congenital and neonatal abnormalities. About 60% are in the head and neck region. The proper identification and classification of each type of lesion has a direct impact on the diagnosis and consequently on the treatment of these patients. The Optical Coherence Tomography (OCT) technique obtains high resolution, real time generated images that allow non-invasive visualization of bio tissue structures with a spatial resolution of the order of 10&mu;m, which is 20 times greater than that of an ultrasound, approaching the conventional histological level. That is why it is considered an optical biopsy. In this thesis, images of lesions from 27 patients were obtained using a Doppler OCT system operating at 1325nm, with the purpose of providing images that allowed to evaluate the differences in the architecture of the vessels in different types of lesions located in the cutaneous and mucosa tissue in head and neck regions. The results of this clinical study proved the viability of OCT to distinguish between normal skin and lesioned skin. In vascular malformation, the blood vessels had an average diameter of 114 &mu;m &plusmn; 92 &mu;m and had a mean depth of 304,3 &mu;m &plusmn; 99 &mu;m. The skin hemangiomas had blood vessels with a mean diameter of 38,6 &mu;m &plusmn; 19 &mu;m and a mean depth of 297,8 &mu;m &plusmn; 132 &mu;m. In the lesions clinically diagnosed as oral mucosa hemangioma (lips), the mean vessel diameter was 45 &mu;m &plusmn; 16 &mu;m and a mean depth of 520,5 &mu;m &plusmn; 189,5 &mu;m. In lip lymphatic malformation, the blood vessels had a mean diameter of 157,5 &mu;m &plusmn; 90 &mu;m and a mean depth of 695 &mu;m &plusmn; 136 &mu;m. In lip vascular malformation, the blood vessels had a mean diameter of 128 &mu;m &plusmn; 63 &mu;m and a mean depth of 689 &mu;m &plusmn; 217 &mu;m. With this information obtained from the Doppler OCT images indicating flow, it was possible to evaluate peculiar differences of the lesion of each patient, promoting the differential diagnosis of vascular lesions of the tissue surface of the skin and oral mucosa.

La tomographie optique diffuse et de fluorescence résolue en temps (TR-TODF) est une méthode qui permet de fournir une information sur les propriétés optiques de diffusion et d’absorption des tissus biologiques. Ce manuscrit de thèse fait l’état de l’art de la méthode et propose des pistes pour reconstruire des images multidimensionnelles 2D/3D des cartes optiques du milieu. L’objectif ultime du projet présenté dans ce document est de concevoir un appareil de mesure (tomographe), éventuellement portatif, pour détecter la présence de tumeurs. Le défi est de pouvoir obtenir des images avec une ré- solution suffisante pour être utilisée en milieu hospitalier à des fins de diagnostic préclinique. Hors le caractère naturellement mal posé du problème inverse rend la tâche complexe. La première partie du document est consacrée à la modélisation du problème physique. En particulier nous nous intéressons à l’approximation de diffusion de l’équation du transfert radiatif dans un milieu quelconque. Dans une deuxième partie, nous traitons le problème du point de vue mathématiques en considérant le problème direct de diffusion couplé avec de la fluorescence pour deux type de mesures: en mode contact et non-contact. Puis nous nous intéressons au problème inverse comme un problème de minimisation d’une fonctionnelle que nous traitons par une méthode de l’adjoint. Enfin et pour finir, la troisième partie du document détaille les différents aspects numérique pour parvenir à un code de reconstruction efficace à l’aide de techniques issues du calcul haute performance. / The Time-Resolved Diffuse Optical Tomography and Fluorescence (TR-DOTF) is a method to obtain optical properties information on diffusion and asbsorption of biological tissues. This Phd manuscript details this method state of the art and highlight the different possible path to reconstruct multidimensionnal 2D/3D images for the optical maps of the turbid medium. The project ultimate goal is to build a measurement instrument (tomograph), eventually portative, in order to detect tumours presence. The challenge is to obtain images with sufficient resolution to be used in medical environment for preclinical diagnosis. However the inverse problem ill-posedness makes the situation more difficult. The first part of this document is devoted to the problem modelization. In particular, we are interested to the diffusion approximation for the radiative transfer equation in a turbid medium. In a second part, we treat this problem from a mathematical point of view considering the diffusion problem coupled with fluorescence for two measurement types: contact and non-contact. Then we focus on the inverse prob- lem as a minimization problem for cost objective function solved by an adjoint method. Last, but not least, the third part of this document details the different numerical aspects involved to achieve an efficient reconstruction code using advanced technics from the high performance computing world.

[pt] Este trabalho propõe um novo algoritmo de segmentação baseado em crescimento de regiões para detecção de nódulos pulmonares em imagens de tomografia computadorizada. Para decidir, em cada iteração, se dois objetos adjacentes são fundidos em um único objeto, o algoritmo de segmentação calcula um índice de heterogeneidade baseada em múltiplos critérios. Entretanto, o algoritmo de segmentação depende de alguns parâmetros os quais foram encontrados utilizando algoritmo genético. Resultados experimentais mostraram que o método é robusto e promissor (chegando a uma sensibilidade de 80,9 % com 0,23 falsos positivos por exame). Além disso, indicam que o método proposto é capaz de fornecer um bom suporte para o diagnóstico do especialista. / [en] This study proposes a novel segmentation algorithm for lung nodules detection in thorax computed tomography (CT). In order to decide, at each iteration, whether two adjacent objects should be merged or not, a region growing procedure calculates a heterogeneity growth based on multiple criteria. However, segmentation algorithm depends on some parameters which were found by genetic algorithm. Results produced by the proposed segmentation were closely consistent with the reference segments provided manually by an expert physician. The detection itself achieved 80,9% sensitivity with 0,23 false positive per slice which indicates that the proposed method is able to provide a good suggestion for the specialist. Results indicate the potential of proposed segmentation method and encourage a further investigation aiming at improving its accuracy.

High-resolution sonar systems are primarily used for ocean floor surveys and port security operations but produce images of limited resolution. In turn, a sonar-specific methodology is required to detect and classify underwater unexploded ordnance (UXO) using the low-resolution sonar data. After researching and reviewing numerous approaches the Multiple Aspect-Fixed Range Template Matching (MAFR-TM) algorithm was developed. The MAFR-TM algorithm is specifically designed to detect and classify a target of high characteristic impedance in an environment that contains similar shaped objects of low characteristic impedance. MAFR-TM is tested against a tank and field data set collected by the Sound Metrics Corp. DIDSON US300. This thesis document proves the MAFR-TM can detect, classify, orient, and locate a target in the sector-scan sonar images. This paper focuses on the MAFR-TM algorithm and its results. / by Lisa Nicole Brisson. / Thesis (M.S.C.S.)--Florida Atlantic University, 2010. / Includes bibliography. / Electronic reproduction. Boca Raton, Fla., 2010. Mode of access: World Wide Web.