No description available.

Tratamentos braquiterápicos são comumente realizados conforme o relatório da American Association of Physicists in Medicine (AAPM), Task Group report TG-43U1, o qual define o formalismo para cálculo de dose absorvida na água e não considera a composição dos materiais, densidades, dimensões do paciente e o efeito dos aplicadores. Estes efeitos podem ser significantes, conforme descrito pelo recente relatório da AAPM, Task Group report TG- 186, que define diretrizes para que sistemas de planejamento modernos, capazes de considerar as complexidades descritas acima, sejam implementados. Esta tese tem como objetivo contribuir para o aumento da exatidão dos planejamentos de tratamento braquiterápicos, seguindo as recomendações do TG-186 e indo além do mesmo. Um software foi desenvolvido para integrar planejamentos de tratamento e simulações pelo método de Monte Carlo (MC); modelos acurados, CAD-Mesh, foram utilizados para representar aplicadores braquiterápicos; Grandezas utilizadas para reportar dose absorvida, Dw,m (dose para água no meio) e Dm,m (dose para o meio no meio), foram calculadas para um tratamento de cabeça e pescoço, considerando a teoria para pequenas (SCT small cavity theory) e grandes cavidades (LCT large cavity theory); a componente da dose em razão do movimento da fonte foi avaliada para tratamentos de próstata e ginecológicos. Perfis de velocidade obtidos na literatura foram utilizados; medidas de velocidade de uma fonte braquiterapica foram realizadas com uma câmera de alta taxa de aquisição. Cálculos de dose obtidos usando MC (incluindo a composição e densidade dos tecidos, ar e o aplicador) mostram sobredoses de aproximadamente 5% dentro do volume alvo, em um tratamento ginecológico, quando comparados aos resultados obtidos com um meio homogêneo de água. Por sua vez, subdoses de aproximadamente 5% foram observadas ao considerar a composição dos tecidos e regiões com ar em um tratamento intersticial de braço. Um aplicador cilíndrico oco resultou na sobredose observada no caso ginecológico, ressaltando a necessidade de modelos acurados para representar os aplicadores. Os modelos CAD-Mesh utilizados incluem um aplicador Fletcher-Williamson, com blindagem, e um balão deformável para irradiação de mama. Os resultados obtidos com estes modelos são equivalentes aos obtidos com modelos geométricos convencionais. Este recurso pode ser conveniente para aplicadores complexos e/ou quando o projeto dos aplicadores for disponibilizado pelo fabricante. Cálculos de dose, com a composição real dos tecidos humanos, podem apresentar diferenças significativas em razão da grandeza adotada. Diferenças entre Dm,m e Dw,m (SCT ou LCT) chegam a 14% em razão da composição do osso. A metodologia adotada (SCT ou LCT) resulta em diferenças de até 28% para o osso e 36% para os dentes. A componente de dose de trânsito também pode levar a diferenças significativas, uma vez que baixas velocidades ou movimentos uniformemente acelerados foram descritos na literatura. Considerando a pior condição e sem incluir nenhuma correção no tempo de parada, a dose de trânsito pode chegar a 3% da dose prescrita para um caso ginecológico, com 4 cateteres, e até 11.1% da dose prescrita para um tratamento de próstata, com 16 cateteres. A dose de trânsito para a fonte avaliada (velocidade obtida experimentalmente) não é uniformemente distribuída e pode levar a sub ou sobredoses de até 1.4% das doses comumente prescritas (310 Gy). Os tópicos estudados são relevantes para tratamentos braquiterápicos e podem contribuir para o aumento de sua acurácia. Os efeitos estudados podem ser avaliados com o uso do software, associado a um código MC, desenvolvido. / Brachytherapy treatments are commonly performed using the American Association of Physicists in Medicine (AAPM) Task Group report TG-43U1 absorbed dose to water formalism, which neglects human tissue densities, material compositions, body interfaces, body shape and dose perturbations from applicators. The significance of these effects has been described by the AAPM Task Group report TG-186 in published guidelines towards the implementation of Treatment Planning Systems (TPS) which can take into account the above mentioned complexities. This departure from the water kernel based dose calculation approach requires relevant scientific efforts in several fields. This thesis aims to improve brachytherapy treatment planning accuracy following TG-186 recommendations and going beyond it. A software has been developed to integrate clinical treatment plans with Monte Carlo (MC) simulations; high fidelity CAD-Mesh geometry was employed to improve brachytherapy applicators modelling; different dose report quantities, Dw,m (dose to water in medium) and Dm,m (dose to medium in medium), were obtained for a head and neck case using small cavity theory (SCT) and large cavity theory (LCT); the dose component due to the source moving within the patient was evaluated for gynecological and prostate clinical cases using speed profiles from the literature. Moreover, source speed measurements were performed using a high speed camera. Dose calculations using MC showed overdosing around 5% within the target volume for a gynecological case comparing results obtained including tissue, air and applicator effects against a homogeneous water phantom. On the other hand, the same comparison showed underdosing around 5% when including tissue and air composition for an interstitial arm case. A hollow cylinder applicator was responsible for the overdosing observed for the gynecological case highlighting the importance of accurate applicator modelling. The evaluated CAD-Mesh applicators models included a Fletcher- Williamson shielded applicator and a deformable balloon used for accelerated partial breast irradiation. Results obtained were equivalent to ones obtained with conventional constructive solid geometry and may be convenient for complex applicators and/or when manufacturer CAD models are available. Differences between Dm,m and Dw,m (SCT or LCT) are up to 14% for bone in a evaluated head and neck case. The approach (SCT or LCT) leads to differences up to 28% for bone and 36% for teeth. Differences can also be significant due to the source movement since some speed profiles from literature show low source speeds or uniform accelerated movements. Considering the worst case scenario and without include any dwell time correction, the transit dose can reach 3% of the prescribed dose in a gynecological case with 4 catheters and up to 11.1% when comparing the average prostate dose for a case with 16 catheters. The transit dose for a high speed (measured with a video camera) source is not uniformly distributed leading to over and underdosing, which is within 1.4% for commonly prescribed doses (310 Gy). The main subjects evaluated in this thesis are relevant for brachytherapy treatment planning and can improve treatment accuracy. Many of the issues described in here can be assessed with the software, coupled with a MC code, developed in this work.

Tratamentos braquiterápicos são comumente realizados conforme o relatório da American Association of Physicists in Medicine (AAPM), Task Group report TG-43U1, o qual define o formalismo para cálculo de dose absorvida na água e não considera a composição dos materiais, densidades, dimensões do paciente e o efeito dos aplicadores. Estes efeitos podem ser significantes, conforme descrito pelo recente relatório da AAPM, Task Group report TG- 186, que define diretrizes para que sistemas de planejamento modernos, capazes de considerar as complexidades descritas acima, sejam implementados. Esta tese tem como objetivo contribuir para o aumento da exatidão dos planejamentos de tratamento braquiterápicos, seguindo as recomendações do TG-186 e indo além do mesmo. Um software foi desenvolvido para integrar planejamentos de tratamento e simulações pelo método de Monte Carlo (MC); modelos acurados, CAD-Mesh, foram utilizados para representar aplicadores braquiterápicos; Grandezas utilizadas para reportar dose absorvida, Dw,m (dose para água no meio) e Dm,m (dose para o meio no meio), foram calculadas para um tratamento de cabeça e pescoço, considerando a teoria para pequenas (SCT small cavity theory) e grandes cavidades (LCT large cavity theory); a componente da dose em razão do movimento da fonte foi avaliada para tratamentos de próstata e ginecológicos. Perfis de velocidade obtidos na literatura foram utilizados; medidas de velocidade de uma fonte braquiterapica foram realizadas com uma câmera de alta taxa de aquisição. Cálculos de dose obtidos usando MC (incluindo a composição e densidade dos tecidos, ar e o aplicador) mostram sobredoses de aproximadamente 5% dentro do volume alvo, em um tratamento ginecológico, quando comparados aos resultados obtidos com um meio homogêneo de água. Por sua vez, subdoses de aproximadamente 5% foram observadas ao considerar a composição dos tecidos e regiões com ar em um tratamento intersticial de braço. Um aplicador cilíndrico oco resultou na sobredose observada no caso ginecológico, ressaltando a necessidade de modelos acurados para representar os aplicadores. Os modelos CAD-Mesh utilizados incluem um aplicador Fletcher-Williamson, com blindagem, e um balão deformável para irradiação de mama. Os resultados obtidos com estes modelos são equivalentes aos obtidos com modelos geométricos convencionais. Este recurso pode ser conveniente para aplicadores complexos e/ou quando o projeto dos aplicadores for disponibilizado pelo fabricante. Cálculos de dose, com a composição real dos tecidos humanos, podem apresentar diferenças significativas em razão da grandeza adotada. Diferenças entre Dm,m e Dw,m (SCT ou LCT) chegam a 14% em razão da composição do osso. A metodologia adotada (SCT ou LCT) resulta em diferenças de até 28% para o osso e 36% para os dentes. A componente de dose de trânsito também pode levar a diferenças significativas, uma vez que baixas velocidades ou movimentos uniformemente acelerados foram descritos na literatura. Considerando a pior condição e sem incluir nenhuma correção no tempo de parada, a dose de trânsito pode chegar a 3% da dose prescrita para um caso ginecológico, com 4 cateteres, e até 11.1% da dose prescrita para um tratamento de próstata, com 16 cateteres. A dose de trânsito para a fonte avaliada (velocidade obtida experimentalmente) não é uniformemente distribuída e pode levar a sub ou sobredoses de até 1.4% das doses comumente prescritas (310 Gy). Os tópicos estudados são relevantes para tratamentos braquiterápicos e podem contribuir para o aumento de sua acurácia. Os efeitos estudados podem ser avaliados com o uso do software, associado a um código MC, desenvolvido. / Brachytherapy treatments are commonly performed using the American Association of Physicists in Medicine (AAPM) Task Group report TG-43U1 absorbed dose to water formalism, which neglects human tissue densities, material compositions, body interfaces, body shape and dose perturbations from applicators. The significance of these effects has been described by the AAPM Task Group report TG-186 in published guidelines towards the implementation of Treatment Planning Systems (TPS) which can take into account the above mentioned complexities. This departure from the water kernel based dose calculation approach requires relevant scientific efforts in several fields. This thesis aims to improve brachytherapy treatment planning accuracy following TG-186 recommendations and going beyond it. A software has been developed to integrate clinical treatment plans with Monte Carlo (MC) simulations; high fidelity CAD-Mesh geometry was employed to improve brachytherapy applicators modelling; different dose report quantities, Dw,m (dose to water in medium) and Dm,m (dose to medium in medium), were obtained for a head and neck case using small cavity theory (SCT) and large cavity theory (LCT); the dose component due to the source moving within the patient was evaluated for gynecological and prostate clinical cases using speed profiles from the literature. Moreover, source speed measurements were performed using a high speed camera. Dose calculations using MC showed overdosing around 5% within the target volume for a gynecological case comparing results obtained including tissue, air and applicator effects against a homogeneous water phantom. On the other hand, the same comparison showed underdosing around 5% when including tissue and air composition for an interstitial arm case. A hollow cylinder applicator was responsible for the overdosing observed for the gynecological case highlighting the importance of accurate applicator modelling. The evaluated CAD-Mesh applicators models included a Fletcher- Williamson shielded applicator and a deformable balloon used for accelerated partial breast irradiation. Results obtained were equivalent to ones obtained with conventional constructive solid geometry and may be convenient for complex applicators and/or when manufacturer CAD models are available. Differences between Dm,m and Dw,m (SCT or LCT) are up to 14% for bone in a evaluated head and neck case. The approach (SCT or LCT) leads to differences up to 28% for bone and 36% for teeth. Differences can also be significant due to the source movement since some speed profiles from literature show low source speeds or uniform accelerated movements. Considering the worst case scenario and without include any dwell time correction, the transit dose can reach 3% of the prescribed dose in a gynecological case with 4 catheters and up to 11.1% when comparing the average prostate dose for a case with 16 catheters. The transit dose for a high speed (measured with a video camera) source is not uniformly distributed leading to over and underdosing, which is within 1.4% for commonly prescribed doses (310 Gy). The main subjects evaluated in this thesis are relevant for brachytherapy treatment planning and can improve treatment accuracy. Many of the issues described in here can be assessed with the software, coupled with a MC code, developed in this work.

The purpose of this thesis was to analyse and compare HDR to SDR in relation to the visual components which exists in a picture. The questions asked how big of an impact HDR would have on the visual components in a picture. HDR simulated images of the same scene was analysed based on the visual component to compare in what ways they would be altered. The definition of a visual component is based on Bruce Blocks book The Visual Story. The result of the analysis exposed that HDR alters the current visual components and its sub components. Another conclusion was that the method used has it flaws and would need more research, the complication being the hardware needed to properly view HDR content. / Syftet med detta examensarbete var att analysera och jämföra HDR mot SDR i relation till de visuella komponenter som finns i en bild. Frågan som ställdes var hur stor inverkan HDR har på bildens visuella komponenter. HDR-simulerade bilder från samma miljö analyserades baserat på de visuella komponenter för att se hur de ändrades. Defintionen av vad som räknas som en visuell komponent kommer från Bruce Blocks bok The Visual Story. Resultatet av analysen var att HDR förändrar de nuvarande komponenterna och underkomponenter. En till slutsats var att metoden har sina svagheter och skulle behöva studeras mer ingående, svårigheterna är att ett specifikt hårdvarustöd krävs för att kunna visa HDR korrekt.

Title: Geophysical survey for deep geothermal energy exploitation at the site Nova Paka Author: Magda Karousová Supervisor: Prof. RNDr. Miloš Karous, DrSc. Supervisor's e-mail address: karous@geonika.com The topic of this thesis is a literature search of geological, geophysical, and geothermal information about wider surroundings of the area of Nova Paka and a complete geophysical survey focused on mapping of deeper fault zones and verification of Permo-Carboniferous sediments thickness. The area has been preliminarily assessed as promising for the use of deep geothermal energy. It belongs to the south-western part of Krkonoše Piedmont Basin. The bedrock of the basin is formed by Proterozoic metamorphites. Heating of surrounding rocks by intrusions of local volcanics during the Younger Tertiary is considered important for geothermal applications. The cover of metamorphic bedrock by sediments may indicate accumulation of the heat from the bedrock at relatively shallow depths of several hundred meters. Geophysical measurements, namely reflection and refraction seismics, gravimetry and resistivity profiling identified two major fault zones of east-west and northwest-southeast direction. The interpretation of the data obtained by geophysical sounding methods, namely seismic reflection and vertical...

A number of QA procedures have been developed for Breast Brachytherapy treatments, yet none guarantee accurate dose delivery or allow conformation of the actual source position leading to errors sometimes going unnoticed. The objective of this study is to track the exact path the HDR source would follow in real time. The exit radiation of the HDR source was used to image a well defined matrix of markers. The images were acquired using FPD and were processed to obtain projection coordinates while an x-ray calibration image was processed to obtain marker coordinates. Each marker along with its projection represents a line in 3D. A mathematical solution for the ‘near-intersection’ of two 3D lines was implemented and used to determine the ‘true’ 3D source position. A matrix with N markers will produce N*(N-1)/2 points of intersection and their mean will result in a more accurate source position. This study has proved that the accuracy of source position detection using a FPD is sub-millimeter.

Orientador: João Roberto Gomes de Faria / Banca: Paulo Sérgio Sacarazzato / Banca: Luiz Gonzaga Campos Porto / Resumo: O conforto visual no ambiente de trabalho também é alvo de estudo da ergonomia, e com o auxílio da análise de imagens digitais - HDR (High Dynamic Range) é possível ter um estudo mais rápido e barato contribuindo assim para o bem estar dos trabalhadores, já que uma boa iluminação tornará o ambiente de trabalho mais produtivo e prazeroso. Tem-se por objetivo especificamente, estudar formas de caracterizar o ambiente em relação à possibilidade da ocorrência de ofuscamento a partir de imagens HDR e viabilizar o uso de câmeras fotográficas digitais de uso amador, com lentes padrão, no desenvolvimento de uma sistemática e da correspondente instrumentação para avaliar índices de ofuscamento empregando imagens HDR compostas a partir de fotos obtidas com tais câmeras / Abstract: The visual comfort in the work environment also is white of study of the ergonomics, and with the aid of the analysis of digital images - HDR (High Dynamic Range) is possible to have a cheap study faster e thus constributing it welfare of the workers, science a good illumination will became the environment of more productive and pleasant work. It is had specifically for objective, to study forms to characterize the environment in relation a possibility da ofuscamento occurrence to leave of images HDR and to make possible the use of digital cameras of amoteur use, with leses standard, no development of a systematics and da corresponding instrumentation to evaluate ofuscamento indices being used composed images HDR to leave of photos gotten with such cameras / Mestre

High Efficiency Video Coding (HEVC/H.265) est la dernière norme de compression vidéo, finalisée en Janvier 20 13. Son extension scalable, SHVC, a été publiée en Octobre 2014 et supporte la scalabilité spatiale, en gamut de couleur (CGS) et même en norme de compression (AVC vers HEVC). SHVC peut être utilisée pour l'introduction de nouveaux services, notamment grâce à la rétrocompatibilité qu'elle apporte par la couche de base (BL) et qui est complétée par une couche d'amélioration (BL+EL) qui apporte les nouveaux services. De plus, SHVC apporte des gains en débit significatifs par rapport à l'encodage dit simulcast (l'encodage HEVC séparés). SHVC est considérée par DVB pour accompagner l'introduction de services UHD et est déjà incluse dans la norme ATSC-3.0. Dans ce contexte, l'objectif de la thèse est la conception de stratégies de régulation de débit pour les codeurs HEVC/SHVC lors de l'introduction de nouveaux services UHD. Premièrement, nous avons étudié l'approche p-domaine qui modélise linéairement le nombre coefficient non-nuls dans les résidus transformés et quantifiés avec le débit, et qui permet de réaliser des régulations de débit peu complexes. Après validation du modèle, nous avons conçu un premier algorithme de contrôle de débit au niveau bloc en utilisant cette approche. Pour chaque bloc et son débit cible associé, notre méthode estime de façon précise le paramètre de quantification (QP) optimal à partir des blocs voisins, en limitant l'erreur de débit sous les 4%. Puis, nous avons proposé un modèle d'estimation déterministe du p-domaine qui évite l'utilisation de tables de correspondance et atteignant une précision d'estimation supérieure à90%. Deuxièmement, nous avons investigué l'impact du ratio de débit entre les couches d'un codeur SHVC sur ses performances de compression, pour la scalabilité spatiale, CGS et SOR vers HDR. En se basant sur les résultats de cette étude, nous avons élaborés un algorithme de régulation de débit adaptatif. La première approche proposée optimise les gains de codage en choisissant dynamiquement le ratio de débit optimal dans un intervalle prédéterminé et fixe lors de l'encodage. Cette première méthode a montré un gain de codage significatif de 4.25% par rapport à une approche à ratio fixe. Cette méthode a été ensuite améliorée en lui ajoutant des contraintes de qualité et de débit sur chaque couche, au lieu de considérer un in tervalle fixe. Ce second algorithme a été testé sur le cas de diffusion de programme HD/UHD ct de déploiement de se1vices UHDI-P1 vers UHD 1-P2 (cas d'usage DVB), où elle permet des gains de 7.51% ct 8.30% respectivement. Enfin, le multiplexage statistique de programmes scalable a été introduit et brièvement investigué. Nous avons proposé une première approche qui ajuste conjointement le débit global attribué à chaque programme ainsi que le ratio de débit, de façon à optimiser les performances de codage. De plus, la méthode proposée lisse les variations et l'homogénéité de la qualité parmi les programmes. Cette méthode a été appliquée à une base de données contenant des flux pré-encodés. La méthode permet dans ce cas une réduction du surcoût de la scalabilité de 11.01% à 7.65% comparé à l'encodage a débit et ratio fixe, tout en apportant une excellente précision et une variation de qualité limitée. / High Efficiency Video Coding (HEVC/H.265) is the latest video coding standard, finalized in Janua1y 2013 as the successor of Advanced Video Coding (AVC/H.264). Its scalable extension, called SHVC was released in October 2014 and enables spatial, bitdepth, color-gamut (CGS) and even standard scalability. SHVC is a good candidate for introducing new services thanks to backward compatibility features with legacy HEVC receivers through the base-layer (BL) stream and next generation ones through the BL+EL (enhancement layer). In addition, SHVC saves substantial bitrate with respect to simulcast coding (independent coding of layers) and is also considered by DVB for UHD introduction and included in ATSC-3 .0. In this context, the work of this thesis aims at designing efficient rate-control strategies for HEVC and its scalable extension SHVC in the context of new UHD formats introduction. First, we have investigated the p-domain approach which consists in linking the number of non-zero transfonned and quantized residual coefficients with the bitrate, in a linear way, to achieve straightforward rate-control. After validating it in the context of HEVC and SHVC codings, we have developed an innovative Coding Tree Unit (CTU)-level rate-control algorithm using the p-domain. For each CTU and its associated targeted bit rate, our method accurately estimates the most appropriate quantization parameter (QP) based on neighborhood indicators, with a bit rate error below 4%. Then, we have proposed a deterministic way of estimating the p-domain model which avoids the implementation of look-up tables. The proposed method enables accurate model estimation over 90%. Second, we have explored the impact of the bitrate ratio between layers on the SHVC performance for the spatial, CGS and SDR-to-HDR scalability. Based on statistical observations, we have built an adaptive rate control algorithms (ARC). We have first proposed an ARC scheme which optimizes coding performance by selecting the optimal ratio into a fixed ratio inte1val, under a global bitrate instruction (BL+EL). This method is adaptive and considers the content and the type of scalability. This first approach enables a coding gain of 4.25% compared to fixed-ratio encoding. Then, this method has been enhanced with quality and bandwidth constraints in each layer instead of considering a fixed interval. This second method has been tested on hybrid delivery of HD/UHD services and backward compatible SHVC encoding of UHDI -PI /UHDI -P2 services (DVB use-case) where it enables significant coding gains of 7.51% and 8.30%, respectively. Finally, the statistical multiplexing of SHVC programs has been investigated. We have proposed a first approach which adjusts both the global bit rate to allocate in each program and the ratio between BL and EL to optimize the coding performance. In addition, the proposed method smooths the quality variations and enforces the quality homogeneity between programs. This method has been applied to a database containing pre-encoded bitstreams and enables an overhead reduction from 11.01% to 7.65% compared to constant bitrate encoding, while maintaining a good accuracy and an acceptable quality variations among programs.

This thesis describes experimental work performed (1998-2001) during the author's involvement with the Brachytherapy group at the Peter MacCallum Cancer Centre (PMCC), where he was employed by its Department of Physical Sciences and subsequent modeling and analytical studies. When PMCC added HDR brachytherapy to its radiation therapy practice, an existing operating suite was considered the ideal location for such procedures to be carried out. The integration of brachytherapy into the theatre environment was considered logical due to the relatively invasive nature of brachytherapy techniques and the availability of medical equipment. This thesis contains the detailed study of three key Research Questions involved in clinical aspects relating to quality assurance of an HDR brachytherapy practice. An investigative chapter is dedicated to the pursuit of each of the Research Questions. The first question asked… Is the novel approach to using modular shielding combined with time and distance constraints adequately optimized during HDR brachytherapy? In order to establish optimal clinical practices, this project evaluates the effectiveness of additional shielding added to the modular shielding system without modification of the previously determined time and distance constraints for PMCC staff, other patients, and member of the public. The DOSXYZnrc user code for the EGSnrc Monte Carlo radiation transport code has been used to model exposure pathways to strategic locations used for measurement in and around the operating theatre suite. Modeling allowed exposure pathways to various areas with the facility to be tested without the need to use real sources. The second Research Question asked… How well is dose anisotropy characterized in the near field range of the clinic's HDR 192Ir source? This study experimentally investigated the anisotropy of dose around a 192Ir HDR source in a water phantom using MOSFETs as relative dosimeters. In addition, modeling using the DOSRZnrc user code for the EGSnrc Monte Carlo radiation transport code was performed to provide a complete dose distribution consistent with the MOSFET measurements. Measurements performed for radial distances from 5 to 30 mm extend the range of measurements to 5 mm which has not been previously reported for this source construction. The third Research Question is aimed at the patient level. Is the dose delivered to in vivo dosimeters, located within critical anatomical structures near the prostate, within acceptable clinical tolerance for a large group of HDR prostate patients? An in vivo dosimetry technique employing TLDs to experimentally measure doses delivered to the urethra and rectum during HDR prostate brachytherapy was investigated. Urethral and rectal in vivo measurements for 56 patients have been performed in the initial fraction of four-fraction brachytherapy boost. In the absence of comparable in vivo data, the following local corrective action level was initially proposed: more than 50% of the prostatic urethra receiving a dose 10% beyond the urethral tolerance. The level for investigative action is considered from the analyses of dose differences between measured data and TPS calculation.

This habilitation introduces the question-learn-test-feedback pattern that resulted from a series of investigations in the domains of software testing and modular software construction. The contributions to these fields are of various natures, but are all fundamentally related through two major assumptions: software construction paradigms have to constantly evolve in order to deal with the increasingly complex requirements that software-intensive systems have to meet; a tester's perspective can support this evolution through the development of effective testing techniques and new empirical knowledge about these paradigms. Abstraction, modularity and separation of concerns have been advocated as key factors for rigorous software engineering for a long time. These principles have been incarnated by various software construction paradigms such as object-oriented programming and design, model-driven development and aspect-oriented modeling. These paradigms evolve in order to deal with the increasing number of heterogeneous requirements, the large number of variations and the need for adaptation that software-intensive systems have to integrate. The work presented here is about the integration of effective testing techniques in these paradigms and how this led us towards a more precise understanding of these paradigms. A major discovery in our work is that we could follow a systematic pattern when investigating these paradigms to integrate error detection capabilities. First, we have to question these paradigms about the new assumptions they introduce on software systems. When answering these questions we can can perform the following actions: learn through rigorous evaluation of hypotheses about these paradigms; test software systems developed in these new paradigms; provide feedback to the paradigms in the form of new construction techniques that improve testability. We capture these four facets for the investigation of software construction paradigms in the question-learn-test-feedback pattern (QLTF). This habilitation reports on investigations in three software construction techniques: object-oriented programming and design, aspect-oriented programming, model transformations. Each investigation is synthesized around the question-learn-test-feedback pattern.