As fístulas arteriovenosas da região do seio cavernoso constituem as fístulas carotidocavernosas que podem ser diretas ou indiretas. As indiretas são raras, a sua sintomatologia é variada e o tratamento é controverso. Este estudo compreendeu a análise prospectiva de 44 pacientes portadores de fístulas carotidocavernosas indiretas (FCCI) no período de 01 de janeiro de 1994 e 31 de janeiro de 2004, 42 com etiologia espontânea e dois pacientes com etiologia traumática, sendo estes analisados separadamente. Doze (12) pacientes foram submetidos à conduta expectante e orientados a realizar manobras de compressão carótido-jugular. O tratamento endovascular foi realizado por via arterial, venosa ou combinação dos dois, num total de 30 pacientes. Considerando o grupo inteiro, ocorreu trombose espontânea em aproximadamente 24% dos pacientes. Os sintomas e o aspecto angiográfico após o tratamento evoluíram com melhora ou cura em 100% dos casos, com oclusão completa das FCCI em 63,3%, a grande maioria destes submetidos a apenas um procedimento. Além dos acessos venosos tradicionais aos seios cavernosos, vias de acesso alternativas através da veia oftálmica superior foram realizadas por punção percutânea de veia facial, veia supratroclear ou veia frontal. O material embolizante mais utilizado foi o adesivo tissular líquido, \"cola\", isoladamente ou em conjunto com outros materiais. Houve complicações transitórias em 13,3% dos pacientes tratados e nenhuma complicação permanente foi observada, o que demonstrou a baixa morbidade deste procedimento / The arteriovenous fistulas of the cavernous sinus (CS) region constitute the carotid-cavernous fistula, which can be direct or indirect. The indirect type is quite rare, its clinical features is very inespecific and its treatment modalities controversial. Forty-four patients with indirect carotid-cavernous fistulas (ICCF) were studied in a prospective manner between January 1994 to January 2004, 42 with spontaneous etiology and 2 with traumatic etiology, being these analyzed separately. Twelve (12) patients were submitted to a expectant management and instructed to perform carotid-jugular compression. Endovascular treatment was accomplished by arterial approach, vein approach or combination of both, in a total of 30 patients. Considering the entire group, spontaneous thrombosis was observed in approximately 24%. Symptoms and the angiographic features after endovascular treatment improved or disappeared in 100% of the cases, with total obliteration in 63.3%, most of them submitted to just one procedure. Despite the traditional venous routes to the CS, alternative accesses through the superior ophthalmic vein (SOV) were accomplished by percutaneous puncture of the facial, supratrochlear or frontal vein. Liquid adhesive (glue) was the most often embolic material used isolated or with other materials. No permanent complication was observed and only 13,3% of the patients treated cursed with transitory complications, what demonstrated the low morbidity of this procedure

Hepatocellular carcinoma is a primary liver cancer.Among treatment options for hepatocellular carcinoma, Yttrium-90 radioembolization is a promising transarterial therapy.This thesis investigates potential clinical applications of radioembolization in the treatment of the hepatocellular carcinoma (techniques of radiation segmentectomy and radiation lobectomy), and adress the problematic of the response Assessment after radioembolization. / Doctorat en Sciences médicales (Médecine) / info:eu-repo/semantics/nonPublished

Abstract Endovascular aneurysm repair (EVAR) is a minimally invasive alternative to open surgery to exclude an abdominal aortic aneurysm from the circulation to avert a rupture. The aim of this thesis was to evaluate the early and mid-term results of EVAR using the Zenith® stent-graft (Cook Inc, Bloomington, IN, USA) in asymptomatic and symptomatic abdominal aortic aneurysm (AAA) patients in three Finnish university hospitals. Furthermore, the aim was to study whether preoperative embolization of the inferior mesenteric artery (IMA) before EVAR decreases the incidence of a type II endoleak or has an effect on the aneurysm sac shrinkage. Finally, the results after secondary interventions for a type II endoleak were evaluated. Two hundred six patients underwent elective endovascular repair of an intact AAA. The use of the Zenith® stent-graft was associated with good early and mid-term results. The thirty-day mortality rate (2.9%) was in accordance with other EVAR studies. Only one late aneurysm-related death occurred in this series, whilst no patients died of a late aneurysm rupture. No stent-graft migrations or fractures were observed. Endoleak, defined as persistent blood flow outside the graft and within the aneurysm sac, remains a long-term problem with EVAR. The overall endoleak incidence was 34.6%. A type II endoleak (retrograde perfusion via aortic side branches) occurred in 52 patients (25.4%). EVAR was performed for 14 patients with a symptomatic, unruptured AAA. The median delay from admission to intervention was 4 days. EVAR of a symptomatic, unruptured AAA was associated with a favourable outcome even in patients with a very high operative risk. There were no perioperative deaths. Altogether forty patients treated at Kuopio University Hospital had a patent IMA on preoperative computed tomography (CT) and were treated successfully with coil embolization before EVAR. Thirty-nine patients who underwent EVAR at Oulu University Hospital without preoperative embolization of a patent IMA served as a control group. Preoperative coil embolization of the IMA significantly reduced the incidence of type II endoleaks after EVAR, but the present study failed to show any influence on late postoperative aneurysm sac shrinkage. Overall, 14 patients underwent a secondary intervention to repair the type II endoleak. Ten patients had transarterial embolization and four patients had translumbar embolization. The results were unsatisfactory; clinical success after the first secondary intervention was achieved in only two patients in the transarterial embolization group and three patients in the translumbar embolization group. These results seem to favour direct translumbar embolization rather than transarterial embolization. In conclusion, EVAR with the Zenith® stent-graft is effective in excluding AAAs from the circulation and is associated with good mid-term results.

Portal hypertension (PHT) is a condition with serious complications, such as variceal bleeding, refractory ascites and bowel ischemia. The cause of PHT may be pre-, intra- or post-hepatic. Initial treatment is pressure-reducing drugs and the treatment of acute symptoms. Ten patients presented with severe abdominal pain and acute portomesenteric venous thrombosis. Their response to systemic anticoagulation was insufficient. Treatment with primary continuous thrombolysis by a transhepatic or transjugular approach in four patients resulted in major complications, incomplete recanalization and a 75% survival rate. Treatment with repeated transjugular thrombectomy (TT) combined with the creation of a transjugular intrahepatic portosystemic shunt (TIPS) achieved near complete recanalization, prompt symptom relief and 100% survival in five patients treated with this method as the primary intervention. In one patient, treated with TT and TIPS secondary to surgical thrombectomy and bowel resection, the outcome was fatal. Nineteen patients with portal vein thrombosis presented with acute or threatening variceal bleeding or refractory ascites. TIPS was feasible in 16 of the 18 patients in whom it was attempted and symptom relief was achieved in the majority of them. In 14 patients with Budd-Chiari syndrome, 13 patients were treated with TIPS, four of them after previous liver vein angioplasty. The 5-year transplantation-free survival rate was 100% in patients treated with primary TIPS. In 131 patients with variceal bleeding treated with TIPS, the survival at 12 months in patients with and without cirrhosis was 70% and 100% respectively and in accordance with previous studies. A high Child-Pugh score prior to TIPS and severe HE within 12 months after TIPS was related to an increased mortality. The occurrence of HE after TIPS did not correlate with the PSG after TIPS. Re-bleeding within 12 months after TIPS occurred in 10 patients and was associated with TIPS dysfunction. In conclusion, endovascular intervention, mainly TIPS, seems to be safe and effective for treating patients with complications of PHT, regardless of the underlying cause of disease and site of venous blood flow obstruction. HE may occur more frequently after TIPS than medical and endoscopic treatment, but is often mild and easily treated. In selected patients with PHT, TIPS may improve survival.

L’optimisation du guidage en radiologie interventionnelle implique de simplifier les gestes, d’améliorer la qualité d’images et leur précision tout en réduisant l’exposition aux rayons X du personnel soignants et des patients et l’usage du produit de contraste iodé, néphrotoxique. Elle implique un travail sur chacune des quatre étapes fondamentales de chaque intervention que sont : « voir », « atteindre », « intervenir » et « contrôler » le succès de l’intervention et l’absence de complication.L’étape « voir » la cible a fait l’objet de 3 études : la première, animale, portait sur l’amélioration de la qualité d’images de cone-beam computed tomography (CBCT) et la seconde, clinique, sur la précision de la segmentation des tumeurs hépatiques mise en évidence sur les images de CBCT lors de chimioembolisations (CHE). Et enfin, la troisième, clinique, sur le repérage des artères épigastriques inférieures profondes en réalité augmentée avant lambeau.L’étape « atteindre » la cible a fait l’objet de 5 études. Deux études ont validé la précision du ciblage des tumeurs hépatiques lors des CHE utilisant des logiciels de détection automatique des artères nourricières et de perfusion virtuelle du territoire tumoral à traiter sur les images de CBCT lors de CHE. Trois autres études portaient sur les cartographies 3D par la technique de fusion d’images 3D projetées sur la fluoroscopie 2D. Les deux premières premières ont montré que l’usage de la fusion d’images lors des traitements endovasculaires des anévrismes complexes de l’aorte abdominale utilisant les images d’angioscanner ou d’angioIRM préopératoires permettant de réduire voire de s’affranchir de l’usage de produit de contraste iodé dans la prise en charge de ces pathologies. Le développement de la fusion d’images notamment lors du placement de shunt porto-systémique trans-hépatique par voie trans-jugulaire a permis de faciliter ces interventions tout en limitant l’usage des rayons X dans une troisième étude.La troisième étape « interventir » a fait l’objet de trois études. Une étude, fantôme, a montré la possibilité d’obternir une imagerie sélective et quantitative des structures iodées par l’imagerie spectrale des microsphères chargées d’huile iodée. Une seconde étude préclinique animale a attesté la radioopacité de ces mêmes microsphères chargées lorsqu’elles s’accumulent dans les capillaires ou néovaisseaux et sont le reflet de la nécessité de combiner une recherche commune sur le couple imagerie/matériel. Et enfin l’étude sur la rentabilité des biospies des tumeurs guidées par l’image qui a permis de mettre fin à des dogmes.La quatrième étape « contrôler » le succès des traitements comprend deux études. La première, est une revue de la littérature et a permis de proposer une standardisation de l’usage du CBCT lors des CHE et la seconde a évalué différents facteurs prédicteurs de réponse tumorale en IRM après CHE avec un intérêt particulier porté sur l’usage de nouveaux critères d’analyse 3D (vRECIST et qEASL).Le guidage en radiologie et notamment par le CBCT est encore confronté à la limitation majeure qu’est d’utiliser des rayons X et des produits de contraste iodé mais également à un champ de vue limité et à une qualité d’images sensibles aux artéfacts. D’ou le développement d’autre moyen de guidage comme l’IRM, éléctromagnétisme, fibroptique ou encore la réalité augmentée. Ces derniers éléments font partis de projets de recherche en cours ou à venir. / Optimizing image guidance in interventional radiology involves simplifying procedures, improving image quality and accuracy while reducing X-rays exposure, and the use of iodinated contrast media, which is nephrotoxic. It involves to improve each of the four fundamental steps of each intervention: "see", "reach", "intervene" and "control" the success of the intervention and the absence of complication.To "see" the target was the subject of 3 studies: the first study, was an animal study focused on improving the quality of images of cone-beam computed tomography (CBCT). The second study was a clinical study about the precision of the tumor segmentation of on CBCT images during chemoembolization (TACE). And finally, the third study was a clinical study and described the identification of the lower deep epigastric arteries in augmented reality before flap.To "reach" the target was the subject of 5 studies. Two studies investigated the accuracy of hepatic tumor targeting in TACE using automatic feeding detection and virtual tumor perfusion software on CBCT images. Three other studies were focused on 3D roadmap using the image fusion technique overlaid onto 2D fluoroscopy. The two first studies were about endovascular treatments of complex abdominal aortic aneurysms using images fusion based on angiography of computed tomography or IRM to reduce or even eliminate the use of iodinated contrast agent in the management of such disease. The development of image fusion, particularly during trans-hepatic porto-systemic shunt placement, was shown to facilitate these interventions while limiting the use of X-rays published in to a thrid study.To "intervene" step included three studies. A phantom study showed that spectral imaging displayed selective and quantitative images of iodine content of iodine-loaded microspheres. A preclinical study demonstrated their visibility on x-ray based imaging when they accumulated in capillaries or neovessels and reflected the need to combine a common search for the couple imaging / biomaterials. And finally the study on the rentability of the image guided biospies of tumors which allowed to end dogmas.To "control" the success of treatments included two studies. The first study is a review of the literature and allowed a standardization of the use of the CBCT during the TACE and the second evaluated the factors predictors of tumor response on MRI images after TACE by the use of new 3D criteria. Image guidance with the use of CBCT, is still confronted with the major limitation of the use of X-rays and iodinated contrast media, but also to a limited field of view and a sentitive images quality to artefacts. Hence the development of other image guidance types such as MRI, electromagnetism, fibroptic or augmented reality are on its way. These last elements are part of a research projects in progress or to come.

During catheterized x-ray interventions the patient and medical staff is exposed to scatter radiation, as a consequence of tissue interactions. Ionizing radiation for medical purpose is potentially dangerous and can cause malignancy, skin damage and more. Studies have suggested an increase in the prevalence of eye lens cataract, thyroid cancer and left sided brain tumors in doctors. Therefore, it is mandatory to reduce the radiation dose in medicine, a principle known as ALARA (as low as Reasonably Achievable). Lead aprons, collars and shieldings are safety precautions to protect the team in the operating room. The x-ray equipment and surgical techniques are constantly evolving and the interventions become more complex which may increase the x-ray dose. Although x-ray imaging is required in interventional procedures endeavors of reducing radiation exposure to staff is of high interest. There is a need to increase the awareness about scatter radiation and radiation protection efforts are gaining momentum. Initiative to train a dose reducing behavior by education and awareness are key documents within the European Union’s guidelines on Radiation protection. The aims of this thesis were to create a 3D model for representation of real-time exposure and accumulated scatter radiation to staff performing interventional x-ray procedures and identify parameters that affect the scatter radiation. Extensive measurements were made with real time dosimeters while irradiating an anthropomorphic phantom. For five lateral C-arm projections, 68 - 80 data points each were used to measure scatter dose distribution around the patient. In the typical operator position, the effect of craniocaudal projection angle, patient size, field size, image detector height and pulse rate on scatter radiation dose was also investigated. It was possible to create a 3D model from interpolated measurement data that can generate dose rate with promising results. Six out of eight modelled doses deviated +/- 26.6 % from the validation cases. A model that delivers relative dose is an intuitive approach in education for interventional x-ray radiation safety. The staff position in relation to the x-ray source and the patient size have a significant correlation to the dose rate. Additional measurements are needed to ensure the reliability of the model. This work completes the effect of scatter radiation distribution around the patient table, which is not yet evaluated as thoroughly by other authors.

INTRODUÇÃO: A angioplastia carotídea com stent (ACS) sob proteção cerebral é opção terapêutica em pacientes com estenose carotídea. Existe o risco de embolia apesar da utilização do filtro e as modificações na perfusão cerebral após tratamento da estenose carotídea não são claras. O propósito deste estudo é avaliar, após ACS sob proteção cerebral por filtros, modificações nas seqüências de RM de difusão (DWI) e perfusão (PWI), correlacionando-as com os aspectos técnicos da ACS, com as características da estenose e com dados demográficos dos pacientes. MÉTODO: Trinta e seis pacientes portadores de estenose carotídea com idade média de 72,08 anos foram submetidos a exame de RM um dia antes e até 72 horas após a ACS com filtro de proteção. Todos os pacientes eram assintomáticos após a ACS. Áreas de restrição na DWI após a ACS foram correlacionadas com aspectos demográficos, com aspectos da técnica de angioplastia e com a presença de infartos prévios por RM. Os parâmetros CBV volume sanguíneo cerebral, MTT tempo de trânsito médio e TTP tempo para o pico são empregados para análise por PWI. RESULTADOS: Na DWI, 18 de 36 (50,00%) pacientes apresentaram novos focos (NF) de restrição na DWI após ACS. Todos os NF foram clinicamente silenciosos (100%). Estes NF eram localizados em território cerebral nutrido pela artéria carótida submetida à ACS em 77,19% e menores que 10 mm em 91,53%. Os NF em território cerebral não irrigado pela artéria carótida submetida à angioplastia correspondiam a 22,81% destes. A presença de infartos cerebrais prévios na RM foi o único fator com influência no aparecimento de NF (p=0,037). Fatores demográficos e aspectos relacionados com a técnica de angioplastia não tiveram importância na gênese dos NF. Na PWI foi observada melhora nos parâmetros temporais TTP (p<0,001) e MTT (p=0,019) quando comparados de forma normalizada em relação ao território contralateral. CONCLUSÃO: Os novos focos de restrição na DWI após ACS (NF) foram mais comuns no território ipsilateral (77,19%), no entanto houve NF no território contralateral à ACS (22,81%), possivelmente, associados ao cateterismo diagnóstico. Os NF, na sua maioria, são de pequeno diâmetro (<10 mm em 91,53%). Melhora precoce na PWI, observada nos dados normalizados, foi demonstrada nos parâmetros temporais (TTP e MTT). / INTRODUCTION: Carotid angioplasty with stent (CAS) under cerebral protection is a therapeutic option in patients with carotid stenosis. There is a risk of embolism even with a filter, and changes in cerebral perfusion after treatment are not clearly understood. The purpose of this study was to evaluate changes in diffusion- (DWI) and perfusion- (PWI) weighted magnetic resonance imaging (MRI) sequences correlating them with the technical aspects of CAS, stenosis characteristics and patient demographic data. METHODS: Thirty-six carotid stenosis patients with an mean age of 72.08 years were submitted to MRI exam one day before and up to 72 hours after CAS with filter protection. All patients were asymptomatic after CAS. Areas of restriction on DWI were correlated to demographic aspects, technique of angioplasty as well the presence of previous stroke by MRI. The parameters, CBV - cerebral blood volume; MTT - mean transit time, and TTP- time to peak, are used for PWI analysis. RESULTS: Eighteen of the 36 patients (50.00%) presented new focus (NF) of restriction by DWI after CAS. All new focus were clinically silent. The NF were located in the cerebral area fed by the carotid artery submitted to CAS in 77.19% and smaller than 10mm in 91.53%. NF in cerebral area not irrigated by carotid artery submitted by angioplasty correspond to 22,81 %. The presence of previous ischemic lesion on MRI was the only factor which influenced the appearance of NF (p=0.037). Demographic factors and aspects related to angioplasty technique had no importance on NF genesis. Improvement in PWI timing parameters - TTP (p<0.001) and MTT (p=0.019) were observed in relation to the contralateral territory (normalized data). CONCLUSION: The restriction NF in the DWI after CAS are more common in the ipsilateral territory (77.19%), however there were some NF in the contralateral territory to the CAS (22.81%), possibly associated with diagnostic catheterization. Most of the NF were small in diameter (<10mm in 91.53%). Short-term improvement in PWI were demonstrated by normalized timing parameters (TTP and MTT).

Thermoablationsverfahren werden in der klinischen Routine zur Therapie bei Patienten mit primären Lebertumoren und Lebermetastasen eingesetzt, bei denen von einer Operation abgesehen wird. Die laserinduzierte Thermotherapie (LITT) ist ein minimal invasives radiologisches Verfahren zur perkutanen Tumorablation. Mit der Magnetresonanztomographie (MRT) am Hochfeldtomographen (1.5 Tesla) steht eine radiologische Methode mit der Option einer präzisen Prozesskontrolle der Thermoablation und einer suffizienten Erfolgskontrolle zur Verfügung. Ziel dieser Arbeit war es, im Tierexperiment die Anwendung der LITT zur Ablation von Pankreasgewebe zu erproben, das Potenzial der MRT für die Prozesskontrolle der LITT am Pankreas zu bestimmen und neue MRT-Sequenzen mit neuen Kontrastmitteln für die Optimierung der Erfolgskontrolle zu erforschen. Die LITT am Pankreas im Rahmen einer Pilotstudie an 15 Läuferschweinen war perkutan komplikationsarm durchführbar, generalisierte Pankreatitiden oder Blutungen traten nicht auf. Die qualitative Prozessbeobachtung mittels thermosensitiver Sequenzen zeigte eine hohe Übereinstimmung zwischen magnetresonanztomographisch dokumentierten Thermoeffekten und histopathologisch verifizierten thermisch induzierten Nekrosen. Die Untersuchung und invasive Kalibrierung verschiedener Messmethoden in vivo zur quantitativen MRT-Thermometrie ergab Vorteile für den Einsatz der Protonenresonanzfrequenz-Methode. Zur Optimierung der Erfolgskontrolle nach LITT von Lebergewebe im Tierexperiment sowie klinisch bei Lebermetastasen wurden die MRT-Kontrastmittel Gadomesoporphyrin, Eisenoxid und Gadobutrol erprobt. Mittels Spätaufnahmen 6 - 18 h post injectionem wurden mit Gadobutrol thermisch induzierte Nekrosen präzise visualisiert. / Thermoablation of primary liver tumors and liver metastases is widely used in patients without surgical options. The laser-induced thermotherapy (LITT) is a minimal invasive radiologic procedure for percutaneous tumor ablation. With high field magnetic resonance imaging at 1.5, monitoring of thermoablation and visualization of thermal induced ablation zones can be performed precisely. Aim of this work was to investigate the feasibility of MR-guided LITT of pancreatic tissue and to optimise the contrast between thermal induced lesions, residual tumor and normal tissue after LITT procedure. MR-guided LITT was feasible in 15 female pigs, generalized pancreatitis or bleeding did not occur. MR monitoring by thermosensitive sequences precisely visualized thermal induced ablation zones verified by histopathologic examination. Best results of MR thermometry (thermo-mapping) were obtained by proton resonance frequency method. Gadolinum- mesoporphyrine, superparamagnetic iron oxides (SPIO) and gadobutrol were used to optimise ablation control. Late enhanced imaging 6 - 18 hours after injection of gadobutrol precisely visualized thermal induced necrosis. In conclusion, percutaneous MR guided LITT of pancreatic tissue of female pigs was feasible and monitoring of thermoablation could be performed accurately. In contrast to other imaging methods, MR using new contrast agents enables accurate visualization of thermal induced necrosis.

Trinta e cinco pacientes com malformações arteriovenosas raquimedulares (MAVR) foram tratados por via endovascular. Foram analisados os tipos de MAVR, bem como os aspectos clínicos e angiográficos pré e pós-tratamento endovascular. As MAVR mais freqüentes foram as fístulas arteriovenosas durais (FAD), com 17 pacientes, seguidas das malformações arteriovenosas intramedulares (MAI), fístulas arteriovenosas perimedulares (FAVP) e malformações arteriovenosas metaméricas (MAVM). A avaliação do resultado do tratamento foi dividida em sucesso técnico e clínico, sendo que o primeiro foi subdividido em sucesso técnico do procedimento e do tratamento. O sucesso técnico do procedimento foi caracterizado como oclusão total da MAVR ao final de um procedimento endovascular. O sucesso técnico do tratamento foi caracterizado pela oclusão definitiva da MAVR, independentemente do número de procedimentos. A recidiva foi definida como o reaparecimento da MAVR, identificada na angiografia durante o seguimento dos pacientes, independentemente do quadro clínico apresentado. A recidiva implicava em insucesso do procedimento, mas não necessariamente do tratamento. O sucesso clínico do tratamento foi considerado como a melhora parcial ou total dos sintomas após um período mínimo de acompanhamento de 6 meses. Nas FAD, houve predomínio do sexo masculino e da faixa etária acima dos 36 anos. O tratamento endovascular das FAD mostrou sucesso do procedimento na maioria dos casos, porém com recidiva da lesão em 56,7% dos mesmos. Quando usadas partículas, observou-se recidiva em 73,3% dos procedimentos, porém com sucesso do tratamento em 66,7 dos pacientes. Quando utilizado o adesivo tissular (\"cola\"), houve recidiva em 11,1% dos procedimentos e sucesso do tratamento em 88,9% dos pacientes. A melhora clínica foi obtida em 64,7% dos pacientes. Dos 10 pacientes tratados com MAI, a faixa etária mais freqüente foi dos 19 aos 35 anos de idade e predomínio do sexo masculino. O tratamento endovascular das MAI mostrou sucesso do procedimento em 53,3% e recidiva da lesão em 6,7% dos pacientes. O sucesso do procedimento também foi superior quando usada cola, atingindo 63,6% dos procedimentos e apenas 25,0% com partículas. O sucesso do tratamento foi de 80,0% e a melhora clínica foi observada em 40,0% dos pacientes. Somente 1 paciente com MAVM foi tratado. O tratamento objetivou a embolização parcial da lesão para alívio dos sintomas, resultado que foi obtido com sucesso. Nas FAVP houve predomínio do sexo masculino e a faixa etária mais acometida foi dos 19 aos 35 anos. O tratamento endovascular das FAVP mostrou sucesso do procedimento em 75,0% dos pacientes e ausência de recidiva. O sucesso do tratamento foi de 85,7%, com melhora clínica em todos os pacientes. O nosso estudo mostrou que as MAVR podem ser tratadas de maneira eficiente e segura através da embolização. O seguimento dos pacientes tratados com cola mostrou menor índice de recidiva do que aqueles tratados com partículas. Após a embolização, houve melhora significativa no padrão da marcha nos pacientes com FAD e MAI / Endovascular treatment was used in 35 patients with spinal cord arteriovenous malformations (SCAVM). SCAVM types, along with clinical and angiographic features of pre-op and post-op endovascular treatments were analyzed. Seventeen patients had dural arteriovenous fístulas (DAVF), being this type the most frequent SCAVM, followed by intramedullary arteriovenous malformations (IAVM), perimedullary arteriovenous fístulas (PAVF) and metameric arteriovenous malformations (MAVM). Evaluation of treatment results was divided into technical and clinical success, being the first subdivided into technical success of the procedure and treatment. The procedural\'s technical success was characterized by total SCAVM occlusion at the end of an endovascular procedure. Technical success of the treatment was characterized by definitive occlusion of the SCAVM, independent of the number of procedures. Recurrence was defined as reappearance of SCAVM, identified on angiograms during patient follow up, independent of the observed clinical situation. Recurrence implied in an unsuccessful procedure but not necessarily in treatment failure. Treatment\'s clinical success was considered partial or total symptom improvement after a minimum 6 month period follow up. Male and age over 36 years was preponderate in DAVF. Endovascular treatment in DAVF had procedural success in most cases, but with lesion recurrence in 56,7% of these patients. When particulate embolic agents were used, recurrence occurred in 73,3% of the procedures but with treatment success in 66,7% of these patients. When liquid tissue adhesive was used, there was 11,1% of recurrence and 88,9% in treatment success in these patients. Clinical improvement occurred in 64,7% of these patients. Male predominance and age between 19 to 35 years was more frequent in the 10 treated patients with IAVM. IAVM endovascular treatment was a procedural success in 53,3% and lesion recurrence in 6,7% of these patients. Procedural success was also superior when liquid tissue adhesive was used, up to 63,6% of these procedures and, only 25,0% with particulate embolic agents. Treatment success occurred in 80,0% and clinical improvement seen in 40,0% of these patients. Only 1 patient with MAVM was treated. Aim in treatment was partial lesion embolization to mitigate symptoms, which was successfully achieved. Male predominance and age between 19 to 35 years occurred in PAVF. PAVF endovascular treatment had procedural success in 75,0% and no recurrences in these patients. Treatment success occurred in 85,7% and clinical improvement in all patients. This study demonstrates that SCAVM can be treated efficiently and safely with endovascular embolization. Patient follow up revealed fewer recurrences in those treated with liquid tissue adhesive than in those with particulate embolic agents. After DAVF and IAVM embolization, patients had significant improvement in their gait

近年来，血管类疾病已经成为人类健康的第一杀手。每年有成百上千万人死于血管疾病。血管介入术是一种非常有前景的血管类疾病的治疗手段。血管介入术是一种微创手术，它已经被广泛的用于治疗中风，血管狭窄，血管瘤等疾病。相对于传统的开放式手术，它具有风险低，恢复快，住院时间短等优点。该疗法通常在透视影像的引导下由导管和导线在血管内协同完成手术过程。因为介入术的复杂性和特殊性，作为介入手术医生的必要技能，掌握手术中手眼协同，各种手术器具的使用和复杂细致的手术流程无疑是一个巨大的挑战。因此，迫切地需要一种高效、安全的训练系统。相对于传统的训练方法，基于虚拟现实技术的训练系统是一种非常好的训练手段。 / 为了建立一套高仿真的介入手术训练模拟器，首先，我们要为病人的血管网重建三维模型。我们提出了一种自动的提取中心线的方法，用来从分割好的CTA/MRA体数据中获取病人血管网的中心线。基于改进的平行传递算法，沿着这些中心线，生成了一系列连续的标架。根据这些标架，我们构造了血管的横截面，并在此基础上生成了光滑连续的三维血管模型。 / 其次，作为血管介入术中最基础和最重要的手术器械，我们为导管和导线建立了物理模型。我们提出了一种基于最小势能原理的可变形的模型用于模拟导管和导线对于受力的反应。我们还提出了一个快速并且稳定的多网格算法来保证模拟的真实性和严格的实时交互要求。另外，我们做了几组实验。通过这些实验，验证了多网格算法在稳定性、实时性、模拟的真实性等方面满足了我们对于训练用模拟系统的要求。 / 再次，为了模拟血管栓塞术的手术过程，我们提出了一种模拟线圈填充血管瘤的过程的新方法。通过加总线圈弯曲变形的弹性势能、血管瘤变形的弹性势能以及外力做的功，我们建立了在血管栓塞术的环境下的总势能模型。为了求解这个模型，我们提出了一个基于有限元方法的求解器。从而模拟了线圈在介入医生的操作下慢慢的进入血管瘤，并缠绕起来的过程。 / 另外，我们提出了一个分层圆柱网格模型（LCGM）用于模拟在血管网中血流的运动。这一模型在几何上和拓扑结构上都非常适合我们的应用。我们将血液在血管中的流动近似为一维的层流，并用一组线性等式描述了血管网中流速与血压的关系。通过求解这一线性系统，得到了在分层圆柱网格模型下血流的速度场。依据这个血流的速度场，我们采用平流-扩散模型来模拟造影剂在血管中的传播的过程。 / Vascular diseases have been becoming the number one cause of death worldwide in recent years. Millions of people were killed by vascular diseases each year. An increasingly promising therapy for treating vascular diseases is Vascular Interventional Radiology (VIR). VIR is a minimally invasive surgery (MIS) procedure, which has been widely used to cure stroke, angiostenosis, aneurysm and etc. A low risk, an accelerated recovery and a shorter stay in hospital are important advantages over the traditional vascular surgery. This therapy is performed by a guidewire-catheter combination inside the blood vessels under the guidance of the fluoroscopic imaging. Because of the complexity and particularity of these procedures, it is a great challenge to master hand-eye coordination, instrument manipulation and procedure protocols for each radiologist mandatory. An efficient and safe training system is needed urgently. In contrast to these traditional training methods, virtual reality (VR) based simulation systems is a pretty good surrogate. / In order to build a high fidelity interventional simulator for physician training, firstly, we reconstructed the three dimensional (3D) model for the vascular network of the patients. An method of automatic skeleton extraction was proposed to acquire the centerline of the vascular network from the segmented volume data from CTA/MRA. A series of continuing frames were generated along with the centerline based on improved parallel transporting method. According to these frames we built the crossections of the vessels and further the 3D vascular model with the smooth meshes. / Secondly, as the most basic and important instruments in the VIR procedure, the catheter and guidewire were modeled and simulated physically. We developed a deformable model to simulate complicated behaviors of guidewires and catheters based on the principle of minimum total potential energy. A fast and stable multigrid solver was proposed to ensure both realistic simulation and real time interaction. A series of experiments were conducted to evaluate our multigrid solver in terms of stability, time performance, the capability of simulating catheter behaviors and the realism of catheter deformation. / Thirdly, to simulate the procedure of embolization, we proposed a novel method to simulate the motion of coil and their interactions with the aneurysm. We formulated the total potential energy in the embolization circumstance by summing up the elastic energy deriving from the bending of coils, the potential energy due to the deformation of the aneurysm and the work by the external forces. A novel FEM-based approach was proposed to simulate the deformation of coils. And the motion of coils and their responses to every input from the interventional radiologist can be calculated globally. / Fourthly, we proposed our Layered Cylindrical Gird Model (LCGM) for simulating blood flow in vascular network, which is pretty suitable for sampling the vascular network geometrically and topologically. The blood flow in vessels was regarded as 1D laminar flow and formulated into a set of linear equations based on the Poiseuille law to describe the relationship between the speed of flow and the pressure. Solving those equations, we got the velocity fields in the blood flow. In terms of the velocity fields, an advection-diffusion model was adopted to simulate the propagation of contrast agent with the blood flow. / Finally, all above techniques and procedures were implemented and integrated into a simulation system for training the medical students to acquire the endovascular skill, and an empirical study was also designed based on a typical selective catheteriza- tion procedure to assess the feasibility and effectiveness of the proposed system. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / 最后，我们将所有以上提到的技术和方法集成到模拟系统中用于训练医学院的学生，并使他们获得血管介入术的技能。并且，我们基于一个典型的导管插入术过程，使用经验分析的方法对模拟系统的可用性和效率进行了评估。 / Li, Shun. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 105-116). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts in also Chinese. / Abstract --- p.i / Acknowledgement --- p.vi / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Vascular Modeling --- p.14 / Chapter 2.1 --- Introduction and Related Work --- p.14 / Chapter 2.2 --- Vascular Skeleton Graph Construction --- p.15 / Chapter 2.2.1 --- Chamfer distance transform and Dijkstra's shortest-path algorithm --- p.17 / Chapter 2.2.2 --- End vertices retrieval --- p.19 / Chapter 2.2.3 --- The algorithm of vascular skeleton extraction --- p.21 / Chapter 2.3 --- Vascular Modeling --- p.21 / Chapter 2.3.1 --- Tubular Model --- p.21 / Chapter 2.3.2 --- Bifurcation Model --- p.23 / Chapter 3 --- Catheter Simulation --- p.28 / Chapter 3.1 --- Introduction and Related Works --- p.28 / Chapter 3.2 --- Catheter Simulation --- p.31 / Chapter 3.2.1 --- Kirchhoff Theory of Elastic Rod --- p.32 / Chapter 3.2.2 --- Problem Formulation --- p.34 / Chapter 3.2.3 --- The Multigrid Iterative Solver --- p.38 / Chapter 3.3 --- Collision detection --- p.45 / Chapter 3.4 --- Validation of the Catheter Simulation Method --- p.47 / Chapter 3.4.1 --- Stability --- p.49 / Chapter 3.4.2 --- Time Performance --- p.50 / Chapter 3.4.3 --- Preservation of Curved Tip --- p.51 / Chapter 3.4.4 --- The realism of catheter deformation --- p.53 / Chapter 4 --- Coil Embolization Simulation --- p.59 / Chapter 4.1 --- Introduction and Related Work --- p.59 / Chapter 4.2 --- Methodology --- p.61 / Chapter 4.2.1 --- Total potential energy of a coil --- p.61 / Chapter 4.2.2 --- The FEM-based numeric solver for interactive coil simulation --- p.61 / Chapter 5 --- Angiography Simulation --- p.70 / Chapter 5.1 --- Introduction and related works --- p.70 / Chapter 5.2 --- The Equations of Fluid --- p.72 / Chapter 5.3 --- Layered Cylindrical Gird Model --- p.73 / Chapter 5.4 --- Numerical Method --- p.76 / Chapter 5.4.1 --- Evaluation of the velocity field of blood flow --- p.76 / Chapter 5.4.2 --- Evaluation of the density field --- p.78 / Chapter 5.5 --- Results --- p.81 / Chapter 6 --- System Implementation and Evaluation --- p.84 / Chapter 6.1 --- Introduction and Related Work --- p.84 / Chapter 6.2 --- System Construction --- p.85 / Chapter 6.3 --- Empirical Study of the Training System --- p.89 / Chapter 7 --- Conclusion and Discussion --- p.98 / Chapter 7.1 --- Geometric Modeling of Vasculature --- p.99 / Chapter 7.2 --- Catheterization Simulation --- p.99 / Chapter 7.3 --- Embolization Simulation --- p.100 / Chapter 7.4 --- Angiography Simulation --- p.101 / Chapter 7.5 --- System and Evaluation --- p.102 / Publication List --- p.103 / Bibliography --- p.105