Les tumeurs solides relèvent pour un très grand nombre de cas d’un traitement chirurgical. La réussite thérapeutique dépend de la qualité, de la précision de l’exérèse et de la capacité à « visualiser » les résidus tumoraux. Pour ce faire, différents marqueurs sont disponibles sur le marché, notamment radioactifs. Depuis peu, des émetteurs de positrons spécifiques sont disponibles, relançant l’intérêt des techniques de chirurgies radioguidées. Les objectifs de ces travaux ont été la conception, le développement et l’optimisation d’un détecteur de positons pour la localisation de la tumeur avant exérèse et des amas tumoraux résiduels après exérèse. L’avantage de la détection des positons est leur parcours dans les tissus mous, de quelques millimètres. Leur détection est donc plus précise et de meilleure sensibilité (1 à 3 ordres de grandeurs) que la détection de rayonnements gamma. Elle consiste en une technique d’imagerie de contact sur une profondeur de quelques millimètres. Nous avons donc développé 2 prototypes d’imageurs beta+ composés de photomultiplicateurs en silicone (SiPM). Le SiPM sont des photomultiplicateurs présentant les mêmes caractéristiques que les tubes photomultiplicateurs (PMT), tout en étant sensibles à une faible luminosité et de très petite dimension, ce qui est parfaitement adapté à la problématique. La tête de détection du premier prototype est l’association d’un scintillateur organique sur une matrice de SiPM. Cette configuration permet de réduire le bruit dû à la contamination du signal par des rayonnements gamma. La seconde configuration inclus une méthode de soustraction permettant d’améliorer la discrimination du bruit Gamma, sans l’éliminer totalement. Pour ce faire, un assemblage de deux scintillateurs séparés par un guide de lumière est utilisé. Le signal issu des extrémités du scintillateur sont discriminés par l’analyse des différentes distributions de lumière sur la matrice SiPM. Dans les deux cas, le détecteur a été conceptualisé de façon à positionner la tête de détection dans un boitier mécanique comprenant une électronique de lecture miniaturisée. Différents designs de la sonde Beta+, prenant en compte le matériau et l’épaisseur du scintillateur, la fenêtre de propagation lumineuse et le réflecteur optique ont été simulés avec MonteCarlo et mesurés expérimentalement. Ces paramètres ont été optimisés de façon à ce que la sonde offre les meilleures performances en termes de sensibilité de détection des positons, discrimination du bruit Gamma, résolution et distorsion spatiale et uniformité de réponse. L’influence des différents algorithmes de reconstruction sur les performances spatiales ont été étudiées et une première étude préclinique du détecteur sur fantôme a pu être réalisée avec la première configuration de détecteur. / Extent and accuracy of surgical resection is a crucial step in the therapy of operable solid tumors. The recent availability of specific tumor-seeking agents, positron labeled, renewed the interest for radioguided surgery. The detection of beta particles, due to their short range, allows a more sensitive and accurate tumor localization. Since no mechanical collimation is necessary, it is possible to design probes with a sensitivity increased by one to three orders of magnitude compared to gamma detectors. The beta particle short range also reduces the contamination from distal non-specific radiotracers uptake region, which results in a increased signal-to-noise-ratio. Conversely, beta detection requires sensors to be extremely compact in order to operate in contact with the surveyed tissues in narrow surgical cavities. This thesis takes place in that context. Its aim was to develop an intraoperative positron imaging probe based on the silicon photomultiplier technology (SiPM) and to evaluate its ability to perform in real time tumor localization and post-operative control of the surgical cavity. During this work, two prototypes of intraoperative positron imaging probe were developed. The first detector design is based on the use of a single organic scintillator coupled to an array of SiPMs. This configuration uses a small sensitive volume to reduce the contamination noise coming from the annihilation gamma rays. The second version of the probe implements a subtraction method allowing to improve gamma rejection efficiency. This configuration uses a stack of two scintillators separated by a light guide. The events interacting in the top and the bottom scintillator are discriminated by the analysis of the different light distributions on the SiPM array. Different designs of the positron imaging probes, including scintillator material and thickness, light spreading window and optical reflector, were investigated with Montecarlo simulations and measurements. Their impact on the probes performances were optimized in terms of positron sensitivity, gamma ray rejection efficiency, spatial resolution and bias and uniformity of response. The effect of different reconstruction algorithm on spatial performances was also studied. Finally, the objective of developing an intraoperative probe fully operational in the operating room has been achieved by the design of dedicated miniaturized electronic readouts and mechanical housing. In the last part of my thesis, the evaluation of the single scintillator configuration in a realistic clinical environment was performed with 18F-FDG phantoms. We showed that the low intrinsic sensitivity of this probe to gamma radiations allows to detect tumor volumes as small as 14 mg for uptake properties corresponding to currently available radiotracers and acquisition times compatible with the surgery duration.

Depuis son émergence, la médecine nucléaire ne cesse d’évoluer. A la fois diagnostique et thérapeutique, elle occupe une place importante dans la stratégie médicale moderne. L’imagerie nucléaire consiste après injection au patient d’un radiotraceur, à détecter le rayonnement émis. Elle donne accès quantitativement à la fonctionnalité des organes ou à la localisation de structures cibles telles que des lésions tumorales. Cette imagerie a naturellement intégré les procédures chirurgicales en particulier en cancérologie (en pré et per opératoire). On parle de chirurgie radioguidée. Cette dernière permet de localiser lors du geste chirurgical les structures radiomarquées devant être retirées.Les avancées technologiques au niveau des radiopharmaceutiques et en instrumentation sont à l’origine de nouvelles stratégies de radioguidages pouvant cibler de petites structures. L’imagerie par rayonnement gamma reste la plus répandue et la mieux adaptée. On assiste au développement de dispositifs d’imagerie gamma portables miniaturisés permettant un contrôle visuel en per opératoire. Ces dispositifs sont prometteurs mais doivent être évalués.Un état des lieux des différentes procédures en chirurgie radioguidée et des imageurs utilisés en clinique est réalisé dans cette thèse.Le laboratoire Imagerie et Modélisation en Neurobiologie et Cancérologie (UMR 8165) développe de longue date de nouvelles approches de détection miniaturisée pour les différents types de rayonnement. Parmi celles-ci le prototype de deuxième génération appelé TReCam.Cette mini gamma caméra présente un champ de vue de 4,9 x 4,9 cm2 et intègre des technologies d’imagerie directement issues de la physique des particules. Elle est formée d’un collimateur à trous parallèles, d’un scintillateur continu LaBr3 (Ce) lu par un photomultiplicateur multi-anode et son électronique. Le système d’acquisition donne au chirurgien un affichage en temps réel de l’image radioactive.Ce travail de thèse a consisté également à évaluer la place des imageurs portables en chirurgie radioguidée, en particulier mammaire, à travers l’évaluation de la procédure SNOLL (repérage par marquage ɣ de la tumeur (T) et des ganglions sentinelles (GS)) avec TreCam. Il a reposé sur trois parties.Un premier volet a visé l’optimisation des performances de TReCam pour favoriser la localisation de structures peu radioactives dans des temps d’exposition de l’ordre de la dizaine de secondes. Pour ce faire, différentes stratégies d’optimisation des performances ont été mises en place au niveau du collimateur, de l’électronique et des algorithmes de traitement (dont réseaux de neurones) pour améliorer l’homogénéité de la détection.Le deuxième volet visait à objectiver les performances cliniques de TReCam pour la procédure SNOLL et situer les limites de son exploitation. A l’aide de simulations menées sur la plateforme GATE et modélisant la scène opératoire au plus près de la réalité clinique, nous avons montré que TreCam peut détecter des GS jusqu’à 4,5 cm de profondeur et situé à 4 cm de la T. L’impact du temps de pose n’est pas important. Par contre, le choix de la bonne fenêtre en énergie est primordial.Enfin, le troisième volet concrétise l’ambition interdisciplinaire de cette thèse. Il est consacré à l’évaluation clinique de TReCam à travers l’étude de son apport à la procédure SNOLL mammaire. Cette étude prospective interventionnelle incluant de 47 patientes (22 procédures SNOLL utilisant TReCam aux différents temps de la procédure et 25 procédures SNOLL standard). Les résultats ont montré un intérêt qualitatif à l’utilisation de TReCam en apportant un confort visuel lors de la procédure en complément de la sonde monopixel.Ce travail a montré l’intérêt de tels imageurs en chirurgie radioguidée mais aussi situé leurs limites actuelles. Des efforts de développement doivent être poursuivis tant au niveau des détecteurs qu’au niveau des radiopharmaceutiques utilisés pour le repérage. / Since its emergence in the middle of the twentieth century, nuclear medicine continues to evolve. At the same time diagnostic and therapeutic, it occupies an increasingly important place in the modern medical strategy. Nuclear imaging consists of injecting the patient with a radio-tracer to detect the radiation emitted. It provides quantitative access to the functionality of organs or the location of target structures such as tumor lesions. This imaging has been naturally integrated into surgical procedures, particularly in oncology (preoperatively and then intraoperatively). It is called radio-guided surgery. It makes possible to locate the radioactive target which will be removed during surgery.Technological advances in radiopharmaceutical instrumentation are driving new strategies that can target small structures. Gamma-ray imaging remains the most widespread and the most suitable. We are witnessing the development of miniaturized portable gamma imaging devices that allow visual control during surgery. These devices are promising but need to be evaluated.A short state-of-the-art of the various procedures in radioguided surgery and imagers used clinically is carried out in this thesis.For many years, the Imaging and Modeling in Neurobiology and Oncology Laboratory (UMR 8165) has been developing new miniaturized detection approaches for different types of radiation. Among them is the second-generation prototype called TReCam.This mini gamma camera has a field of view of 4.9 x 4.9 cm2 and integrates imaging technologies directly from particle physics. It consists of a collimator with parallel holes, a continuous scintillator LaBr3 (Ce) read by a multi-anode photomultiplier (PSPMT) and its electronics. The acquisition system gives the surgeon a real-time display of the radioactive image.This thesis work also consisted in evaluating the place of portable imagers in radioguided surgery, in particular mammary surgery, through the evaluation of the SNOLL procedure (identification of the tumor (T) and the sentinel lymph nodes (GS) by γ-labeling) with TreCam. It is based on three parts.The first part aimed at optimizing the performance of TReCam to improve the localization of lowradioactive structures with exposure times of around 10 seconds. To do this, different performance optimization strategies have been implemented in the collimator, electronics and processing algorithms (including neural networks) to improve the homogeneity of the detection. These performances were compared to those of a prototype developed at IMNC and integrating a new generation of photodetectors: the SiPM.The second part aimed to objectify the clinical performance of TReCam in the SNOLL procedure and to situate the limits of its exploitation. Using simulations conducted on the GATE platform and modeling the operating scene closer to clinical reality, we have shown that TreCam can detect GS up to 4.5 cm deep and located 4 cm from the T. Impact of the exposure time is not important. On the other hand, choosing the right energy window is essential.Finally, the third part concretizes the interdisciplinary ambition of this thesis. It is devoted to clinical evaluation of TReCam through the study of its contribution to the SNOLL breast procedure. This prospective interventional study included 47 patients (22 SNOLL procedures using TReCam at different times of the procedure and 25 standard SNOLL procedures). The results showed a qualitative interest in the use of TReCam by bringing a visual comfort during the procedure and must be used in addition to the monopixel probes.This work has shown the interest of such imagers in radioguided surgery but also set their current limits. Development efforts must be pursued at the level of both detectors and radiopharmaceuticals used for tracking.

No description available.

Orientador: Cesar Cabello dos Santos / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Ciencias Medicas / Made available in DSpace on 2018-08-09T08:45:28Z (GMT). No. of bitstreams: 1 Duarte_GiulianoMendes_D.pdf: 5275602 bytes, checksum: 5180518eed1c8045749e430f93e8b706 (MD5) Previous issue date: 2007 / Resumo: Objetivos: Estudar a utilização de cirurgia radioguiada com 99mTc-sestamibi associada à ressonância nuclear magnética (RNM) no câncer de mama. Para isso, foram desenvolvidas duas novas técnicas: uma intitulada Avaliação Intra-peratória Radioguiada das Margens (Radioguided Intraoperative Margins Evaluation -RIME) que usa um radiofármaco intravenoso para auxiliar, através de uma sonda detectora de radiação (probe), a ressecção tumoral; a outra intitulada Fusão da Ressonância Nuclear Magnética e Cintilografia Mamária (Magnetic Resonance Imaging/ Scintimammography fusion ¿ MRI/SM) que procura determinar a extensão tumoral. Dessa forma, esta tese visa avaliar a factibilidade das técnicas, o valor da RIME em obter margens livres e o valor da MRI/SM em determinar o tamanho real do tumor comparando com outros exames. Sujeitos e métodos: Foi realizado um estudo experimental descritivo em 23 mulheres com carcinoma ductal invasivo de mama estádios IIA a IIIA, agendadas para mastectomia. Dois a dez dias antes da cirurgia, as pacientes foram submetidas à cintilografia mamária com 99mTc-sestamibi, na qual foi localizado o tumor e estimado o tempo ideal para iniciar a cirurgia radioguiada através de uma curva de contagem de radiação por tempo. No mesmo dia, 20 pacientes realizaram uma ressonância nuclear agnética com gadolíneo. No momento da cirurgia, a mesma dose de 99mTc-sestamibi foi injetada intravenosamente. Esperou-se o tempo ideal estimado previamente e então foi realizada uma segmentectomia com auxílio do probe, que determinou as margens de ressecção. A ressonância foi usada para avaliar comprometimento da pele, fáscia profunda a outros focos de tumor que foram incluídos na segmentectomia. Isso caracterizou a RIME. Após, todas as pacientes foram submetidas à mastectomia. A egmentectomia e a mama residual foram examinadas histologicamente. Paralelamente, foi desenvolvido um programa de computador em linguagem visual que realizou, de forma semi-automática, a fusão das imagens de ressonância e cintilografia em 20 pacientes (MRI/SM). Foi determinado como tumor na MRI/SM a área de intersecção das imagens, o tumor foi medido e comparado com a medida no exame histológico, ressonância nuclear magnética, mamografia e exame clínico. Para análise estatística foram utilizados os testes exato de Fisher, t de Student, Wilcoxon e regressão linear. Resultados: Em todas as pacientes foi possível realizar as técnicas de RIME e MRI/SM (factibilidade de 100%). A RIME permitiu a ressecção do tumor com margens livres em 19 pacientes, com média de margem de 4,8mm. Em 11 casos não havia doença residual na mama após a segmentectomia. A média do tamanho da doença residual na mama foi 3,6mm e geralmente estava localizado próximo ao leito tumoral (<1,5cm em 10 pacientes). Não houve associação entre presença de doença residual e o tamanho do tumor ou estado das margens. A medida do tumor na MRI/SM teve uma correlação com o exame histológico melhor que a ressonância, a mamografia e exame clínico em todos os diâmetros estudados. Conclusões: As técnicas de RIME e MRI/SM são factíveis. A RIME permitiu auxiliar a ressecção tumoral com margens livres na maioria das vezes. A MRI/SM parece ser melhor que ressonância, mamografia e exame clínico para determinar a extensão tumoral / Abstract: Aims: To study the use of radioguided surgery with 99mTc-sestamibi, associated with magnetic resonance imaging (MRI), for breast cancer evaluation. For this, we developed two techniques: the first was dominated Radioguided Intraoperative Margins Evaluation (RIME), a technique that uses a radiopharmaceutical agent to distinguish normal and cancer tissue with a probe; the second was termed Magnetic Resonance Imaging and Scintimammography fusion (MRI/SM) to determine the extension of cancer. Thus, this study aims to assess the feasibility of techniques, the ability of RIME to obtain free margins and the ability of MRI/SM to determine the real tumor extension, in comparison to other examinations. Subjects and methods: A descriptive experimental study was carried out on 23 women with invasive ductal breast carcinoma, stages IIA to IIIA, programmed for mastectomy. Two to 10 days before the surgery, the patients were submitted to a scintimammography with 99mTc-sestamibi to localize the tumor and to estimate the optimal time to begin radioguided surgery, through a curve of radiation count per time. On the same day, 20 patients realized a MRI with gadolinium. At the moment of the surgery, the same dose of 99mTc-sestamibi was injected intravenously, after a previously-estimated optimal time a segmentectomy was performed with a probe that determined the resection margins. MRI was used to evaluate the compromising of the skin, deep fascia and the other tumor foci that were included in the segmentectomy, characterizing the RIME. All the patients were then submitted to mastectomy. The segmentectomy and residual breast were histopathologically examined. In parallel, software was developed in visual language to perform the MRI/SM fusion in 20 patients. The intersection area between the MRI and scintimammography was determined as tumor in MRI/SM and the tumor was measured and compared with the measurement in the pathological, MRI, mammography and clinical examination. The Fisher¿s exact test, Student¿s t-test, Wilcoxon¿s test and linear regression were used for statistical analysis. Results: The RIME and MRI/SM techniques were successfully performed in all patients (feasibility of 100%). The principal tumor was removed by RIME and provided histologically-free margins in 19 patients (mean margins, 4.8mm). Additionally, 11 patients were without residual disease after segmentectomy. The mean size of residual carcinoma was 3.6mm and generally located near the tumor bed (< 1.5cm in 10 patients). There was no significant association between presence of residual disease and tumor size or margin status. The MRI/SM cancer measurements correlated better with pathology than MRI, mammography and clinical examination in all diameters analyzed. Conclusions: The RIME and MRI/SM techniques are feasible. In general, RIME aids in tumor resection with free margins. MRI/SM seems to be better than MRI, mammography and clinical examination to determine the tumoral extension / Doutorado / Tocoginecologia / Doutor em Tocoginecologia