Made available in DSpace on 2015-03-04T18:57:36Z (GMT). No. of bitstreams: 1 bulant.pdf: 14030619 bytes, checksum: 7bdeb54a02eeadbb90a53496f04a524c (MD5) Previous issue date: 2013-02-26 / This work focuses on tomographic reconstruction and three-dimensional arterial structures segmentation problems, understanding both as fundamentals modules of patient specific-based cardiovascular modeling. Is in this context where the possibility of improving information quality contained in medical images, by means of more accurate reconstruction methods, is of interest because of its impact on segmentation algorithm precision, allowing the construction of more realistic vessel geometrical models. Specifically, in this work, we study the reconstruction from projection inverse problem based on the classic simplified model for x-ray interaction, using the Radon transform, in two and three dimensions with different acquisition geometries. Then, we present the standard methods, which are based on analytic inversion of the Radon transform in the form of back-projection. An iterative reconstruction method based on sensitivity analysis, was studied and adapted to work in a parallel and matrix-free form. In addition, a novel reconstruction method, with foundations given by the theoretical framework of variational formulations, is developed. In view of the characteristics of the methodology, the computational implementation is carried out following a distributed computing paradigm on top of a matrix-free form version of the method. We perform parameter sensitivity analysis for each method, and a quantitative comparative analysis among them, using the well known Shepp-Logan head phantom, is also presented. Some preliminary reconstruction studies with real data are presented. In the context of medical image segmentation, we explore two alternatives for arterial structures segmentation from three dimension images acquired on three-Dimensional Rotational Angiography and Computed Tomography Angiography clinical studies. The first methodology is based on a simple seeded region growing algorithm, combined with thresholding and anisotropic diffusion techniques. The second pipeline uses a Level Set approach, initialized with the colliding fronts algorithm. Now days, Level Set methods variants are gaining popularity for vessel segmentation problems. In this work, we present a qualitative comparison by segmenting a numerical phantom, as well as segmenting arterial structures on different anatomical regions of the head and neck. At last, we present some recommendations for method and parameter choosing, as a function of the anatomical region. / Este trabalho foca-se nos problemas da reconstrução tomográfica e da segmentação de estruturas arteriais em três dimensões, entendendo ambos como blocos fundamentais na modelagem do sistema cardiovascular humano orientada a pacientes específicos. Dentro deste contexto, a possibilidade de melhorar a qualidade das informações contidas nas imagens médicas realizadas em estudos clínicos via métodos de reconstrução mais precisos torna-se atraente porque impacta diretamente na acurácia das técnicas de segmentação, e permite a construção de modelos geométricos mais próximos à realidade. Especificamente, neste trabalho, o problema inverso de reconstrução a partir de projeções é estudado empregando o modelo simplificado clássico, via a transformada de Radon, para duas e três dimensões com diferentes geometrias de amostragem. Logo, os métodos de resolução padrão baseados na inversão analítica desta transformada em forma de back-projection foram estudados. Um método iterativo baseado em análise de sensibilidade, denominado aqui como método baseado no gradiente topológico, foi estudado e adaptado para funcionar com um esquema matrix-free em paralelo. Em seguida, um novo método com fundamentos na teoria de formulações variacionais é desenvolvido. Devido às características do método, o mesmo é implementado computacionalmente empregando paralelismo e um formato matrix-free. Um estudo quantitativo sobre os parâmetros de cada método é fornecido, assim como também um estudo comparativo entre todos eles usando o amplamente difundido fantasma de Shepp-Logan. Por fim, estudos preliminares com projeções de estudos de angiografias rotacionais tridimensionais são apresentados. No contexto da segmentação de imagens médicas, duas alternativas são apresentadas para segmentar estruturas arteriais a partir de imagens tridimensionais provenientes de estudos de angiografia rotacional e tomográfica. A primeira metodologia baseia-se num enfoque simples de crescimento de região, combinado com thresholding e suavizado anisotrópico. A segunda metodologia utiliza um enfoque de Level Set, inicializado por uma técnica denominada colliding fronts. Na atualidade diversas variantes do método de Level Set têm ganhado popularidade na segmentação de estruturas arteriais. Neste trabalho comparam-se qualitativamente ambas alternativas com um fantasma numérico, e na segmentação de estruturas arteriais presentes em imagens médicas localizadas em diversas regiões anatômicas. Por fim, recomendações sobre o método mais conveniente a ser utilizado em cada região junto com valores para os parâmetros são fornecidos.

Η CT αγγειογραφία εγκεφάλου (CTA) είναι μέθοδος καθιερωμένη για την διερεύνηση και την θεραπεία των ενδοκρανιακών ανευρυσμάτων. Σκοπός της μελέτης μας ήταν η σύγκριση των ευρημάτων της ψηφιακής αγγειογραφίας (DSA) με αυτά της CTA και με τα χειρουργικά ευρήματα σε ασθενείς με οξεία υπαραχνοειδή αιμορραγία, καθώς επίσης και η αξιολόγηση της κλινικής χρησιμότητας της μεθόδου. Κατά την διάρκεια τριών ετών, 82 ασθενείς προσήλθαν με κλινική εικόνα και σημειολογία συμβατή με υπαραχνοειδή αιμορραγία. Η CTA έγινε αμέσως μετά την απλή CT, ενώ η DSA εντός των πρώτων 48 ωρών από την εισαγωγή. Όλα τα ανευρύσματα που ευρέθησαν με τις δύο μεθόδους υπεβλήθησαν σε χειρουργική αποκατάσταση ή ενδαγγειακό εμβολισμό. Σε όσους ασθενείς βρέθηκε αρνητικό αποτέλεσμα και με τις δύο μεθόδους, έγινε επαναληπτική DSA 15 ημέρες μετά το επεισόδιο με σκοπό την επιβεβαίωση της απουσίας ανευρύσματος. Οι CTA εξετάσεις καθώς και οι κλασσικές αγγειογραφίες μελετήθηκαν από μια ομάδα δύο ακτινολόγων για κάθε τεχνική, οι οποίοι έπρεπε να καταγράψουν την ύπαρξη ή μη ανευρύσματος, να περιγράψουν τα χαρακτηριστικά του και να αξιολογήσουν την μέθοδο. Χειρουργική ή και ενδαγγειακή θεραπεία έγινε σε 45 ασθενείς και ανευρέθησαν 53 ανευρύσματα. Χρησιμοποιώντας την CTA, ευρέθησαν 47 ανευρύσματα σε 42 ασθενείς. Η DSA ανίχνευσε 43 ανευρύσματα σε 39 ασθενείς. Η ευαισθησία της CTA για τον εντοπισμό όλων των ανευρυσμάτων με βάση το χειρουργικό/θεραπευτικό αποτέλεσμα ήταν 88,7%, η ειδικότητα 100%, η θετική προβλεπτική αξία (PPV) 100%, η αρνητική προβλεπτική αξία (NPV) 80,7% και η ακρίβεια 92,3%. Αντίστοιχα, η ευαισθησία της DSA ήταν 87,8%, η ειδικότητα 98%, η PPV 97,7%, η NPV 89,1% και η ακρίβεια 92,9%. Όσον αφορά στα ανευρύσματα ≥3 mm, η CTA είχε ευαισθησία που κυμαινόταν μεταξύ 93,3 έως 100%, ίση με αυτή της DSA. Η CTA εμφάνισε τα ίδια ποσοστά ευαισθησίας με αυτά της DSA σε ανευρύσματα ≥3 mm. Εμφάνισε επίσης 100% ποσοστό ανίχνευσης σε ανευρύσματα της πρόσθιας αναστομωτικής και του διχασμού της μέσης εγκεφαλικής αρτηρίας, ενώ μερικές εντοπίσεις όπως η οπίσθια αναστομωτική αρτηρία παραμένουν προβληματικές. 80 Κατά την διάρκεια της παρούσας μελέτης προσπαθήσαμε να δημιουργήσουμε μια τεχνική προσομοίωσης της διεγχειρητικής εικόνας των ραγέντων ενδοκρανιακών ανευρυσμάτων, με τη χρήση volume rendering techniques σε εικόνες που προκύπτουν από CT αγγειογραφία. Η τρισδιάστατη κατασκευή των εικόνων προέκυψε από την συνεργασία μιας ομάδας αποτελούμενης από τέσσερις ακτινολόγους, έναν νευροχειρουργό και έναν ιατρικό φυσικό. Το αποτέλεσμα αυτής της συνεργασίας ήταν η παραγωγή μιας εικόνας οριοθετημένης στο χώρο, με οδηγά σημεία που εύκολα μπορούσαν να αναπαραχθούν κατά την διάρκεια του χειρουργείου. Οι εικόνες χειρουργικής προσομοίωσης ενός ανευρύσματος είναι πιθανώς χρήσιμο εργαλείο για τον προεγχειρητικό σχεδιασμό των ενδοκρανιακών ανευρυσμάτων. / Cerebral CT angiography is an established method applied to both the detection and treatment planning of intracranial aneurysms. The aim of our study was to compare DSA to CTA findings and with the surgical results mainly in patients with acute SAH and to evaluate the clinical usefulness of CTA. During the last three years, 82 consecutive patients were admitted under clinical symptoms and signs suggestive of harbouring an intracranial aneurysm. CT angiography performed immediately afterwards the plain CT, while DSA was performed within the first 48 hours of admission. All aneurysms detected, were confirmed during surgery or endovascular embolization. Repeat DSA was performed in all patients having both the initial CTA and the DSA 15 days after the onset of symptoms negative. CT angiograms and conventional angiographies were studied by a consensus of two radiologists for each technique, who performed aneurysm detection, morphological features characterization and evaluation of the technique. Surgical or/and endovascular treatment was performed in 45 patients and 53 aneurysms were confirmed. Using 3D-CT angiography we detected 47 aneurysms in 42 patients. Conventional angiography depicted 43 aneurysms in 39 patients. The sensitivity of CTA for the detection of all aneurysms versus surgery was 88.7%, the specificity 100%, the positive predictive value (PPV) 100%, the negative predictive value (NPV) 80.7% and the accuracy 92.3%. Consequently, the sensitivity of DSA was 87.8%, the specificity 98%, the PPV 97.7%, the NPV 89.1% and the accuracy 92.9%. Considering the aneurysms ≥ 3 mm, CTA showed a sensitivity ranging from 93.3% to 100%, equal to that of DSA. Cerebral CT angiography has an equal sensitivity to DSA in the detection of intracranial aneurysms greater than 3 mm. It has also 100% detection rate in AcoA and MCA bifurcation aneurysms, while some locations like posterior communicating artery aneurysms remain problematic. The delineating features of each aneurysm are better depicted with CTA due to 3D visualization. The use of Digital Subtraction Angiography as a diagnostic tool can be limited in equivocal cases. A supplement to the above work is our effort to describe a technique for simulating the surgical view of ruptured intracranial aneurysms, using volume 82 rendering techniques in spiral CT angiography data. The 3D rendered images were assessed by a team consisted of four radiologists, one neurosurgeon and one medical physicist. The resultant ‘surgical view’ image was standardized in space using a three-dimensional coordinate system, which allowed for its reproduction in the operating theatre. The surgical views are easily reproducible and αποτελούν a useful tool for the surgical planning of intracranial aneurysms.