Desenvolvimento de um método semiautomático para geração de ground truths de vídeos

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Class-specific metrics for multidimensional data projection applied to CBIR

Content-based image retrieval is still a challenging issue due to the inherent complexity of images and choice of the most discriminant descriptors. Recent developments in the field have introduced multidimensional projections to burst accuracy in the retrieval process, but many issues such as introduction of pattern recognition tasks and deeper user intervention to assist the process of choosing the most discriminant features still remain unaddressed. In this paper, we present a novel framework to CBIR that combines pattern recognition tasks, class-specific metrics, and multidimensional projection to devise an effective and interactive image retrieval system. User interaction plays an essential role in the computation of the final multidimensional projection from which image retrieval will be attained. Results have shown that the proposed approach outperforms existing methods, turning out to be a very attractive alternative for managing image data sets.

Computerized brain tumor segmentation in magnetic resonance imaging

OBJECTIVE: To propose an automatic brain tumor segmentation system. METHODS: The system used texture characteristics as its main source of information for segmentation. RESULTS: The mean correct match was 94% of correspondence between the segmented areas and ground truth. CONCLUSION: Final results showed that the proposed system was able to find and delimit tumor areas without requiring any user interaction.

Methodological improvement of 3D fluoroscopic analysis for the robust quantification of 3D kinematics of human joints

3D video-fluoroscopy is an accurate but cumbersome technique to estimate natural or prosthetic human joint kinematics. This dissertation proposes innovative methodologies to improve the 3D fluoroscopic analysis reliability and usability. Being based on direct radiographic imaging of the joint, and avoiding soft tissue artefact that limits the accuracy of skin marker based techniques, the fluoroscopic analysis has a potential accuracy of the order of mm/deg or better. It can provide fundamental informations for clinical and methodological applications, but, notwithstanding the number of methodological protocols proposed in the literature, time consuming user interaction is exploited to obtain consistent results. The user-dependency prevented a reliable quantification of the actual accuracy and precision of the methods, and, consequently, slowed down the translation to the clinical practice. The objective of the present work was to speed up this process introducing methodological improvements in the analysis. In the thesis, the fluoroscopic analysis was characterized in depth, in order to evaluate its pros and cons, and to provide reliable solutions to overcome its limitations. To this aim, an analytical approach was followed. The major sources of error were isolated with in-silico preliminary studies as: (a) geometric distortion and calibration errors, (b) 2D images and 3D models resolutions, (c) incorrect contour extraction, (d) bone model symmetries, (e) optimization algorithm limitations, (f) user errors. The effect of each criticality was quantified, and verified with an in-vivo preliminary study on the elbow joint. The dominant source of error was identified in the limited extent of the convergence domain for the local optimization algorithms, which forced the user to manually specify the starting pose for the estimating process. To solve this problem, two different approaches were followed: to increase the optimal pose convergence basin, the local approach used sequential alignments of the 6 degrees of freedom in order of sensitivity, or a geometrical feature-based estimation of the initial conditions for the optimization; the global approach used an unsupervised memetic algorithm to optimally explore the search domain. The performances of the technique were evaluated with a series of in-silico studies and validated in-vitro with a phantom based comparison with a radiostereometric gold-standard. The accuracy of the method is joint-dependent, and for the intact knee joint, the new unsupervised algorithm guaranteed a maximum error lower than 0.5 mm for in-plane translations, 10 mm for out-of-plane translation, and of 3 deg for rotations in a mono-planar setup; and lower than 0.5 mm for translations and 1 deg for rotations in a bi-planar setups. The bi-planar setup is best suited when accurate results are needed, such as for methodological research studies. The mono-planar analysis may be enough for clinical application when the analysis time and cost may be an issue. A further reduction of the user interaction was obtained for prosthetic joints kinematics. A mixed region-growing and level-set segmentation method was proposed and halved the analysis time, delegating the computational burden to the machine. In-silico and in-vivo studies demonstrated that the reliability of the new semiautomatic method was comparable to a user defined manual gold-standard. The improved fluoroscopic analysis was finally applied to a first in-vivo methodological study on the foot kinematics. Preliminary evaluations showed that the presented methodology represents a feasible gold-standard for the validation of skin marker based foot kinematics protocols.

A Fully Automatic Calibration Framework for Navigated Ultrasound Imaging

Navigated ultrasound (US) imaging is used for the intra-operative acquisition of 3D image data during imageguided surgery. The presented approach includes the design of a compact and easy to use US calibration device and its integration into a software application for navigated liver surgery. User interaction during the calibration process is minimized through automatic detection of the calibration process followed by automatic image segmentation, calculation of the calibration transform and validation of the obtained result. This leads to a fast, interaction-free and fully automatic calibration procedure enabling intra-operative

Pedagogical quality in e-learning

The article is concerned with design and use of e-learning technology to develop education qualitatively. The purpose is to develop a framework for a pedagogical evaluation of e-learning technology. The approach is that evaluation and design must be grounded in a learning theoretical approach, and it is argued that it is necessary to make a reflection of technology in relation to activities, learning principles, and a learning theory in order to qualitatively develop education. The article presents three frameworks developed on the basis of cognitivism, radical constructivism and activity theory. Finally, on the basis of the frameworks, the article discusses e-learning technology and, more specifically, design of virtual learning environments and learning objects. It is argued that e-learning technology is not pedagogically neutral, and that it is therefore necessary to focus on design of technology that explicitly supports a certain pedagogical approach. Further, it is argued that design should direct its focus away from organisation of content and towards design of activities.

Visual Fixation for 3D Video Stabilization

Visual fixation is employed by humans and some animals to keep a specific 3D location at the center of the visual gaze. Inspired by this phenomenon in nature, this paper explores the idea to transfer this mechanism to the context of video stabilization for a handheld video camera. A novel approach is presented that stabilizes a video by fixating on automatically extracted 3D target points. This approach is different from existing automatic solutions that stabilize the video by smoothing. To determine the 3D target points, the recorded scene is analyzed with a stateof- the-art structure-from-motion algorithm, which estimates camera motion and reconstructs a 3D point cloud of the static scene objects. Special algorithms are presented that search either virtual or real 3D target points, which back-project close to the center of the image for as long a period of time as possible. The stabilization algorithm then transforms the original images of the sequence so that these 3D target points are kept exactly in the center of the image, which, in case of real 3D target points, produces a perfectly stable result at the image center. Furthermore, different methods of additional user interaction are investigated. It is shown that the stabilization process can easily be controlled and that it can be combined with state-of-theart tracking techniques in order to obtain a powerful image stabilization tool. The approach is evaluated on a variety of videos taken with a hand-held camera in natural scenes.

Dish Detection and Segmentation for Dietary Assessment on Smartphones

Diet-related chronic diseases severely affect personal and global health. However, managing or treating these diseases currently requires long training and high personal involvement to succeed. Computer vision systems could assist with the assessment of diet by detecting and recognizing different foods and their portions in images. We propose novel methods for detecting a dish in an image and segmenting its contents with and without user interaction. All methods were evaluated on a database of over 1600 manually annotated images. The dish detection scored an average of 99% accuracy with a .2s/image run time, while the automatic and semi-automatic dish segmentation methods reached average accuracies of 88% and 91% respectively, with an average run time of .5s/image, outperforming competing solutions.

A multimodal tablet-based interface for designing and reviewing 3D engineering models

The ability to view and interact with 3D models has been happening for a long time. However, vision-based 3D modeling has only seen limited success in applications, as it faces many technical challenges. Hand-held mobile devices have changed the way we interact with virtual reality environments. Their high mobility and technical features, such as inertial sensors, cameras and fast processors, are especially attractive for advancing the state of the art in virtual reality systems. Also, their ubiquity and fast Internet connection open a path to distributed and collaborative development. However, such path has not been fully explored in many domains. VR systems for real world engineering contexts are still difficult to use, especially when geographically dispersed engineering teams need to collaboratively visualize and review 3D CAD models. Another challenge is the ability to rendering these environments at the required interactive rates and with high fidelity. In this document it is presented a virtual reality system mobile for visualization, navigation and reviewing large scale 3D CAD models, held under the CEDAR (Collaborative Engineering Design and Review) project. It’s focused on interaction using different navigation modes. The system uses the mobile device's inertial sensors and camera to allow users to navigate through large scale models. IT professionals, architects, civil engineers and oil industry experts were involved in a qualitative assessment of the CEDAR system, in the form of direct user interaction with the prototypes and audio-recorded interviews about the prototypes. The lessons learned are valuable and are presented on this document. Subsequently it was prepared a quantitative study on the different navigation modes to analyze the best mode to use it in a given situation.

Solución para la delegación de identidad en Sistemas de Gestión de Identidad paneuropeos basada en infraestructuras de certificación y lenguajes formales de asertos de seguridad

La estrategia i2010 de la UE tiene como objetivo garantizar el liderazgo europeo en materia de TIC y poner los beneficios de la Sociedad de la Información al servicio de la economía, la sociedad y la calidad de vida personal, teniendo presente que los éxitos de Europa hasta la fecha se han basado en favorecer la competencia leal en los mercados de las telecomunicaciones y crear un mercado sin fronteras para contenidos y medios de comunicación digitales. En esta línea, la Comisión Europea ha establecido que los distintos estados miembros deben contribuir activamente al desarrollo y uso seguro de los servicios telemáticos entre sus ciudadanos. Más concretamente, atribuye a las Administraciones Públicas, tanto a nivel nacional, regional como local, un papel dinamizador de la Sociedad de la Información que les obliga a ofrecer paulatinamente todos los actos administrativos a los ciudadanos a través de Internet. Como primer paso para el uso seguro de los servicios telemáticos que ofrecen las instituciones públicas se hace preciso dotar a los ciudadanos de una identidad digital que les permita identificarse ante un Proveedor de Servicio o ante otros ciudadanos de manera inequívoca. Por esta razón, la mayoría de países europeos – y otros en el resto del mundo – están promoviendo, sistemas fiables de gestión de identidad electrónica (eIDM), de tal manera que los ciudadanos, las empresas y departamentos gubernamentales (incluso en Estados miembros diferentes) pueden identificar y certificar sus operaciones con precisión, rapidez y sencillez. Sin embargo, la gestión de esta identidad por las Administraciones Públicas supone un importante desafío, acentuado cuando se hace necesaria la interoperabilidad entre Administraciones de diferentes países, puesto que personas y entidades tienen credenciales de identificación diferentes en función de su propio marco jurídico nacional. Consciente del problema, en la Unión Europea se han puesto en marcha una serie de proyectos con el objetivo de conseguir la interoperabilidad de los eIDMs entre las instituciones públicas de diferentes Estados miembros. A pesar de ello, las soluciones adoptadas hasta la fecha son insuficientes porque no prevén todos los posibles casos de interacción del usuario con las instituciones. En concreto, no tienen en cuenta un aspecto muy importante que se ofrece en los distintos sistemas jurídicos nacionales, a saber, la delegación de la identidad, mediante la cual un ciudadano puede autorizar a otro para que actúe en su nombre para acceder a determinados servicios prestados por las instituciones públicas. En esta tesis se realizan un conjunto de aportaciones que dan solución a distintos aspectos de los problemas planteados y que, de forma conjunta, permiten la interoperabilidad y la delegación de identidad en determinados Sistemas de Gestión de Identidad aplicados al entorno de las Administraciones Públicas. En el caso de la delegación, se ha definido un sistema de delegación dinámica de identidad entre dos entidades genéricas que permite solucionar el problema del acceso delegado a los servicios telemáticos ofrecidos por las Administraciones Públicas. La solución propuesta se basa en la generación de un token de delegación, constituido a partir de un Certificado Proxy, que permite a la entidad que delega establecer la delegación de identidad en otra entidad en base a un subconjunto de sus atributos como delegador, estableciendo además, en el propio token de delegación, restricciones en el conjunto de servicios accesibles a la entidad delegada y el tiempo de validez de la delegación. Adicionalmente, se presentan los mecanismos necesarios tanto para poder revocar un token de delegación como para comprobar sin un token de delegación ha sido o no revocado. Para ello se propone una solución para la identificación unívoca de tokens de delegación y la creación de una nueva entidad denominada Autoridad de Revocación de Tokens de Delegación. Entre las características del sistema de delegación propuesto destaca el que es lo suficientemente seguro como para ser utilizado en el entorno de la Administración Pública, que no requiere el uso de mecanismos off‐line para la generación de la delegación y que se puede realizar la delegación de forma instantánea y sin la necesidad de trámites complejos o la participación de un elevado número de entidades. Adicionalmente, el token de delegación propuesto es perfectamente integrable en las infraestructura de clave pública actual lo que hace que, dado que gran parte de las Administraciones Públicas europeas basan sus sistemas de identidad digital en el uso de la PKI y certificados de identidad X.509, la solución pueda ser puesta en marcha en un entorno real sin necesidad de grandes cambios o modificaciones de comportamiento. En lo referente a la interoperabilidad, se realiza un análisis exhaustivo y la correspondiente evaluación de las principales propuestas de Sistemas de Gestión de Identidad orientados a conseguir la interoperabilidad realizadas hasta la fecha en el marco de la Unión Europea y se propone, a alto nivel, una arquitectura de interoperabilidad para la gestión de identidad en las Administraciones Públicas. Dicha arquitectura es lo suficientemente genérica como para poder ser aplicada tanto en el entorno pan‐Europeo como en los entornos nacionales, autonómicos y locales, de tal forma que la interoperabilidad en la gestión de la identidad esté garantizada en todos los niveles de la Administración Pública. Por último, mediante la integración de la solución de delegación dinámica de identidad y la arquitectura de interoperabilidad propuestas se presenta una solución al problema de la delegación en un escenario pan‐Europeo de gestión de identidad, dando lugar a una arquitectura global de interoperabilidad pan‐Europea con soporte a la delegación de identidad. SUMMARY The i2010 European Union Plan aims to ensure European leadership in ICT and to promote the positive contribution that information and communication technologies can make to the economic, social and personal quality of life, bearing in mind that, to date, success in Europe has been based on promoting fair competition in telecommunications markets and on creating a borderless market for contents and digital media. In this line, the European Commission has established that the different member states should contribute actively to the development and secure use of telematic services among their citizens. More specifically, it is attributed to national, regional and local Public Administrations to have a supportive role of the Information Society, requiring them to gradually provide the citizens with Internet‐based access to all administrative procedures acts. As a first step for the secure use of telematic services offered by public institutions, it is necessary to provide the citizens with a digital identity to enable them to identify themselves unequivocally to a Service Provider or to other citizens. For this reason, most European countries ‐ and others in the rest of the world ‐ are promoting reliable systems for managing electronic identity (eIDM), so that citizens, businesses and government departments (even in different Member States) can identify and certify their operations with precision, speed and simplicity. However, the identity management by Public Administrations is a major challenge that becomes more difficult when interoperability between administrations of different countries is needed, due to the fact that individuals and entities have different identification credentials according to their own national legal framework. Aware of the problem, the European Union has launched a series of projects with the aim of achieving interoperability of eIDMs between public institutions of different Member States. However, the solutions adopted to date are insufficient because they do not foresee all possible cases of user interaction with the institutions. In particular, solutions do not take into account a very important aspect that is offered in different national legal systems, namely, the delegation of identity, by which a citizen can authorize another to act on his/her behalf to access certain services provided by public institutions. In this thesis a collection of contributions that provide solution to different aspects of the aforementioned problems are carried out. The solutions, in global, enable interoperability and identity delegation in some of the Identity Management Systems applied to Public Administration environment. In the case of delegation, a dynamic identity delegation system between generic entities is defined. This system makes it possible to solve the problem of delegated access to telematic services offered by Public Administrations. The proposed solution is based on the generation of a piece of information called delegation token. This delegation token, derived from a Proxy Certificate, allows the establishment of identity delegation by an entity that delegates (delegator) in other entity (delegatee) making use of a subset of delegator attributes. It also establishes restrictions on services that can be used by the delegated entity and the expiry date of delegation. In addition to this, the mechanisms necessary to revoke and check the revocation status of a delegation token are presented. To do this, a solution to univocally identify delegation tokens and the creation of a completely new entity, called Token Delegation Revocation Authority, are proposed. The most remarkable characteristics of the proposed delegation system are its security, enough for it to be used in the Public Administration environment, the fact that it does not require off‐line processes in order to generate the delegation, and the possibility of performing the delegation instantaneously and without neither complex processes nor the intervention of a large number of entities. The proposed delegation token can be completely incorporated into current Public Key Infrastructure (PKI). Thus, since most of the European Public Administrations base their digital identity systems on PKI and X.509 identity certificates, the solution can be adopted in a real environment without great changes or performance modifications. Regarding interoperability, an exhaustive analysis and evaluation of most significant proposals on Identity Management Systems that aim to achieve interoperability carried out in the European Union framework until now are performed. A high level identity management interoperability architecture for Public Administrations is also proposed. This architecture is sufficiently generic to be applied to both pan‐European environment and national, regional or local environments, thus interoperability in identity management at all Public Administration levels is guaranteed. Finally, through the integration of the proposed dynamic identity delegation solution and the high level interoperability architecture, a solution to the problem of identity delegation in a pan‐European identity management environment is suggested, leading to a pan‐European global interoperability architecture with identity delegation support.

BioMet®Tools: from modeling and simulation to product design and development

BioMet®Tools is a set of software applications developed for the biometrical characterization of voice in different fields as voice quality evaluation in laryngology, speech therapy and rehabilitation, education of the singing voice, forensic voice analysis in court, emotional detection in voice, secure access to facilities and services, etc. Initially it was conceived as plain research code to estimate the glottal source from voice and obtain the biomechanical parameters of the vocal folds from the spectral density of the estimate. This code grew to what is now the Glottex®Engine package (G®E). Further demands from users in medical and forensic fields instantiated the development of different Graphic User Interfaces (GUI’s) to encapsulate user interaction with the G®E. This required the personalized design of different GUI’s handling the same G®E. In this way development costs and time could be saved. The development model is described in detail leading to commercial production and distribution. Study cases from its application to the field of laryngology and speech therapy are given and discussed.

Segmentation of RV in 4D Cardiac MR Volumes using region-merging graph cuts

Non-invasive quantitative assessment of the right ventricular anatomical and functional parameters is a challenging task. We present a semi-automatic approach for right ventricle (RV) segmentation from 4D MR images in two variants, which differ in the amount of user interaction. The method consists of three main phases: First, foreground and background markers are generated from the user input. Next, an over-segmented region image is obtained applying a watershed transform. Finally, these regions are merged using 4D graph-cuts with an intensity based boundary term. For the first variant the user outlines the inside of the RV wall in a few end-diastole slices, for the second two marker pixels serve as starting point for a statistical atlas application. Results were obtained by blind evaluation on 16 testing 4D MR volumes. They prove our method to be robust against markers location and place it favourably in the ranks of existing approaches.

Dynamic Response Optimization of Vehicles through Efficient Multibody Formulations and Automatic Differentiation Techniques

El diseño y desarrollo de sistemas de suspensión para vehículos se basa cada día más en el diseño por ordenador y en herramientas de análisis por ordenador, las cuales permiten anticipar problemas y resolverlos por adelantado. El comportamiento y las características dinámicas se calculan con precisión, bajo coste, y recursos y tiempos de cálculo reducidos. Sin embargo, existe una componente iterativa en el proceso, que requiere la definición manual de diseños a través de técnicas “prueba y error”. Esta Tesis da un paso hacia el desarrollo de un entorno de simulación eficiente capaz de simular, analizar y evaluar diseños de suspensiones vehiculares, y de mejorarlos hacia la solución optima mediante la modificación de los parámetros de diseño. La modelización mediante sistemas multicuerpo se utiliza aquí para desarrollar un modelo de autocar con 18 grados de libertad, de manera detallada y eficiente. La geometría y demás características de la suspensión se ajustan a las del vehículo real, así como los demás parámetros del modelo. Para simular la dinámica vehicular, se utiliza una formulación multicuerpo moderna y eficiente basada en las ecuaciones de Maggi, a la que se ha incorporado un visor 3D. Así, se consigue simular maniobras vehiculares en tiempos inferiores al tiempo real. Una vez que la dinámica está disponible, los análisis de sensibilidad son cruciales para una optimización robusta y eficiente. Para ello, se presenta una técnica matemática que permite derivar las variables dinámicas dentro de la formulación, de forma algorítmica, general, con la precisión de la maquina, y razonablemente eficiente: la diferenciación automática. Este método propaga las derivadas con respecto a las variables de diseño a través del código informático y con poca intervención del usuario. En contraste con otros enfoques en la bibliografía, generalmente particulares y limitados, se realiza una comparación de librerías, se desarrolla una formulación híbrida directa-automática para el cálculo de sensibilidades, y se presentan varios ejemplos reales. Finalmente, se lleva a cabo la optimización de la respuesta dinámica del vehículo citado. Se analizan cuatro tipos distintos de optimización: identificación de parámetros, optimización de la maniobrabilidad, optimización del confort y optimización multi-objetivo, todos ellos aplicados al diseño del autocar. Además de resultados analíticos y gráficos, se incluyen algunas consideraciones acerca de la eficiencia. En resumen, se mejora el comportamiento dinámico de vehículos por medio de modelos multicuerpo y de técnicas de diferenciación automática y optimización avanzadas, posibilitando un ajuste automático, preciso y eficiente de los parámetros de diseño. ABSTRACT Each day, the design and development of vehicle suspension systems relies more on computer-aided design and computer-aided engineering tools, which allow anticipating the problems and solving them ahead of time. Dynamic behavior and characteristics are thus simulated accurately and inexpensively with moderate computational times and resources. There is, however, an iterative component in the process, which involves the manual definition of designs in a trialand-error manner. This Thesis takes a step towards the development of an efficient simulation framework capable of simulating, analyzing and evaluating vehicle suspension designs, and automatically improving them by varying the design parameters towards the optimal solution. The multibody systems approach is hereby used to model a three-dimensional 18-degrees-of-freedom coach in a comprehensive yet efficient way. The suspension geometry and characteristics resemble the ones from the real vehicle, as do the rest of vehicle parameters. In order to simulate vehicle dynamics, an efficient, state-of-the-art multibody formulation based on Maggi’s equations is employed, and a three-dimensional graphics viewer is developed. As a result, vehicle maneuvers can be simulated faster than real-time. Once the dynamics are ready, a sensitivity analysis is crucial for a robust optimization. To that end, a mathematical technique is introduced, which allows differentiating the dynamic variables within the multibody formulation in a general, algorithmic, accurate to machine precision, and reasonably efficient way: automatic differentiation. This method propagates the derivatives with respect to the design parameters throughout the computer code, with little user interaction. In contrast with other attempts in the literature, mostly not generalpurpose, a benchmarking of libraries is carried out, a hybrid direct-automatic differentiation approach for the computation of sensitivities is developed, and several real-life examples are analyzed. Finally, a design optimization process of the aforementioned vehicle is carried out. Four different types of dynamic response optimization are presented: parameter identification, handling optimization, ride comfort optimization and multi-objective optimization; all of which are applied to the design of the coach example. Together with analytical and visual proof of the results, efficiency considerations are made. In summary, the dynamic behavior of vehicles is improved by using the multibody systems approach, along with advanced differentiation and optimization techniques, enabling an automatic, accurate and efficient tuning of design parameters.

A model to develop a platform for personalised health applications

Antecedentes Europa vive una situación insostenible. Desde el 2008 se han reducido los recursos de los gobiernos a raíz de la crisis económica. El continente Europeo envejece con ritmo constante al punto que se prevé que en 2050 habrá sólo dos trabajadores por jubilado [54]. A esta situación se le añade el aumento de la incidencia de las enfermedades crónicas, relacionadas con el envejecimiento, cuyo coste puede alcanzar el 7% del PIB de un país [51]. Es necesario un cambio de paradigma. Una nueva manera de cuidar de la salud de las personas: sustentable, eficaz y preventiva más que curativa. Algunos estudios abogan por el cuidado personalizado de la salud (pHealth). En este modelo las prácticas médicas son adaptadas e individualizadas al paciente, desde la detección de los factores de riesgo hasta la personalización de los tratamientos basada en la respuesta del individuo [81]. El cuidado personalizado de la salud está asociado a menudo al uso de las tecnologías de la información y comunicación (TICs) que, con su desarrollo exponencial, ofrecen oportunidades interesantes para la mejora de la salud. El cambio de paradigma hacia el pHealth está lentamente ocurriendo, tanto en el ámbito de la investigación como en la industria, pero todavía no de manera significativa. Existen todavía muchas barreras relacionadas a la economía, a la política y la cultura. También existen barreras puramente tecnológicas, como la falta de sistemas de información interoperables [199]. A pesar de que los aspectos de interoperabilidad están evolucionando, todavía hace falta un diseño de referencia especialmente direccionado a la implementación y el despliegue en gran escala de sistemas basados en pHealth. La presente Tesis representa un intento de organizar la disciplina de la aplicación de las TICs al cuidado personalizado de la salud en un modelo de referencia, que permita la creación de plataformas de desarrollo de software para simplificar tareas comunes de desarrollo en este dominio. Preguntas de investigación RQ1 >Es posible definir un modelo, basado en técnicas de ingeniería del software, que represente el dominio del cuidado personalizado de la salud de una forma abstracta y representativa? RQ2 >Es posible construir una plataforma de desarrollo basada en este modelo? RQ3 >Esta plataforma ayuda a los desarrolladores a crear sistemas pHealth complejos e integrados? Métodos Para la descripción del modelo se adoptó el estándar ISO/IEC/IEEE 42010por ser lo suficientemente general y abstracto para el amplio enfoque de esta tesis [25]. El modelo está definido en varias partes: un modelo conceptual, expresado a través de mapas conceptuales que representan las partes interesadas (stakeholders), los artefactos y la información compartida; y escenarios y casos de uso para la descripción de sus funcionalidades. El modelo fue desarrollado de acuerdo a la información obtenida del análisis de la literatura, incluyendo 7 informes industriales y científicos, 9 estándares, 10 artículos en conferencias, 37 artículos en revistas, 25 páginas web y 5 libros. Basándose en el modelo se definieron los requisitos para la creación de la plataforma de desarrollo, enriquecidos por otros requisitos recolectados a través de una encuesta realizada a 11 ingenieros con experiencia en la rama. Para el desarrollo de la plataforma, se adoptó la metodología de integración continua [74] que permitió ejecutar tests automáticos en un servidor y también desplegar aplicaciones en una página web. En cuanto a la metodología utilizada para la validación se adoptó un marco para la formulación de teorías en la ingeniería del software [181]. Esto requiere el desarrollo de modelos y proposiciones que han de ser validados dentro de un ámbito de investigación definido, y que sirvan para guiar al investigador en la búsqueda de la evidencia necesaria para justificarla. La validación del modelo fue desarrollada mediante una encuesta online en tres rondas con un número creciente de invitados. El cuestionario fue enviado a 134 contactos y distribuido en algunos canales públicos como listas de correo y redes sociales. El objetivo era evaluar la legibilidad del modelo, su nivel de cobertura del dominio y su potencial utilidad en el diseño de sistemas derivados. El cuestionario incluía preguntas cuantitativas de tipo Likert y campos para recolección de comentarios. La plataforma de desarrollo fue validada en dos etapas. En la primera etapa se utilizó la plataforma en un experimento a pequeña escala, que consistió en una sesión de entrenamiento de 12 horas en la que 4 desarrolladores tuvieron que desarrollar algunos casos de uso y reunirse en un grupo focal para discutir su uso. La segunda etapa se realizó durante los tests de un proyecto en gran escala llamado HeartCycle [160]. En este proyecto un equipo de diseñadores y programadores desarrollaron tres aplicaciones en el campo de las enfermedades cardio-vasculares. Una de estas aplicaciones fue testeada en un ensayo clínico con pacientes reales. Al analizar el proyecto, el equipo de desarrollo se reunió en un grupo focal para identificar las ventajas y desventajas de la plataforma y su utilidad. Resultados Por lo que concierne el modelo que describe el dominio del pHealth, la parte conceptual incluye una descripción de los roles principales y las preocupaciones de los participantes, un modelo de los artefactos TIC que se usan comúnmente y un modelo para representar los datos típicos que son necesarios formalizar e intercambiar entre sistemas basados en pHealth. El modelo funcional incluye un conjunto de 18 escenarios, repartidos en: punto de vista de la persona asistida, punto de vista del cuidador, punto de vista del desarrollador, punto de vista de los proveedores de tecnologías y punto de vista de las autoridades; y un conjunto de 52 casos de uso repartidos en 6 categorías: actividades de la persona asistida, reacciones del sistema, actividades del cuidador, \engagement" del usuario, actividades del desarrollador y actividades de despliegue. Como resultado del cuestionario de validación del modelo, un total de 65 personas revisó el modelo proporcionando su nivel de acuerdo con las dimensiones evaluadas y un total de 248 comentarios sobre cómo mejorar el modelo. Los conocimientos de los participantes variaban desde la ingeniería del software (70%) hasta las especialidades médicas (15%), con declarado interés en eHealth (24%), mHealth (16%), Ambient Assisted Living (21%), medicina personalizada (5%), sistemas basados en pHealth (15%), informática médica (10%) e ingeniería biomédica (8%) con una media de 7.25_4.99 años de experiencia en estas áreas. Los resultados de la encuesta muestran que los expertos contactados consideran el modelo fácil de leer (media de 1.89_0.79 siendo 1 el valor más favorable y 5 el peor), suficientemente abstracto (1.99_0.88) y formal (2.13_0.77), con una cobertura suficiente del dominio (2.26_0.95), útil para describir el dominio (2.02_0.7) y para generar sistemas más específicos (2_0.75). Los expertos también reportan un interés parcial en utilizar el modelo en su trabajo (2.48_0.91). Gracias a sus comentarios, el modelo fue mejorado y enriquecido con conceptos que faltaban, aunque no se pudo demonstrar su mejora en las dimensiones evaluadas, dada la composición diferente de personas en las tres rondas de evaluación. Desde el modelo, se generó una plataforma de desarrollo llamada \pHealth Patient Platform (pHPP)". La plataforma desarrollada incluye librerías, herramientas de programación y desarrollo, un tutorial y una aplicación de ejemplo. Se definieron cuatro módulos principales de la arquitectura: el Data Collection Engine, que permite abstraer las fuentes de datos como sensores o servicios externos, mapeando los datos a bases de datos u ontologías, y permitiendo interacción basada en eventos; el GUI Engine, que abstrae la interfaz de usuario en un modelo de interacción basado en mensajes; y el Rule Engine, que proporciona a los desarrolladores un medio simple para programar la lógica de la aplicación en forma de reglas \if-then". Después de que la plataforma pHPP fue utilizada durante 5 años en el proyecto HeartCycle, 5 desarrolladores fueron reunidos en un grupo de discusión para analizar y evaluar la plataforma. De estas evaluaciones se concluye que la plataforma fue diseñada para encajar las necesidades de los ingenieros que trabajan en la rama, permitiendo la separación de problemas entre las distintas especialidades, y simplificando algunas tareas de desarrollo como el manejo de datos y la interacción asíncrona. A pesar de ello, se encontraron algunos defectos a causa de la inmadurez de algunas tecnologías empleadas, y la ausencia de algunas herramientas específicas para el dominio como el procesado de datos o algunos protocolos de comunicación relacionados con la salud. Dentro del proyecto HeartCycle la plataforma fue utilizada para el desarrollo de la aplicación \Guided Exercise", un sistema TIC para la rehabilitación de pacientes que han sufrido un infarto del miocardio. El sistema fue testeado en un ensayo clínico randomizado en el cual a 55 pacientes se les dio el sistema para su uso por 21 semanas. De los resultados técnicos del ensayo se puede concluir que, a pesar de algunos errores menores prontamente corregidos durante el estudio, la plataforma es estable y fiable. Conclusiones La investigación llevada a cabo en esta Tesis y los resultados obtenidos proporcionan las respuestas a las tres preguntas de investigación que motivaron este trabajo: RQ1 Se ha desarrollado un modelo para representar el dominio de los sistemas personalizados de salud. La evaluación hecha por los expertos de la rama concluye que el modelo representa el dominio con precisión y con un balance apropiado entre abstracción y detalle. RQ2 Se ha desarrollado, con éxito, una plataforma de desarrollo basada en el modelo. RQ3 Se ha demostrado que la plataforma es capaz de ayudar a los desarrolladores en la creación de software pHealth complejos. Las ventajas de la plataforma han sido demostradas en el ámbito de un proyecto de gran escala, aunque el enfoque genérico adoptado indica que la plataforma podría ofrecer beneficios también en otros contextos. Los resultados de estas evaluaciones ofrecen indicios de que, ambos, el modelo y la plataforma serán buenos candidatos para poderse convertir en una referencia para futuros desarrollos de sistemas pHealth. ABSTRACT Background Europe is living in an unsustainable situation. The economic crisis has been reducing governments' economic resources since 2008 and threatening social and health systems, while the proportion of older people in the European population continues to increase so that it is foreseen that in 2050 there will be only two workers per retiree [54]. To this situation it should be added the rise, strongly related to age, of chronic diseases the burden of which has been estimated to be up to the 7% of a country's gross domestic product [51]. There is a need for a paradigm shift, the need for a new way of caring for people's health, shifting the focus from curing conditions that have arisen to a sustainable and effective approach with the emphasis on prevention. Some advocate the adoption of personalised health care (pHealth), a model where medical practices are tailored to the patient's unique life, from the detection of risk factors to the customization of treatments based on each individual's response [81]. Personalised health is often associated to the use of Information and Communications Technology (ICT), that, with its exponential development, offers interesting opportunities for improving healthcare. The shift towards pHealth is slowly taking place, both in research and in industry, but the change is not significant yet. Many barriers still exist related to economy, politics and culture, while others are purely technological, like the lack of interoperable information systems [199]. Though interoperability aspects are evolving, there is still the need of a reference design, especially tackling implementation and large scale deployment of pHealth systems. This thesis contributes to organizing the subject of ICT systems for personalised health into a reference model that allows for the creation of software development platforms to ease common development issues in the domain. Research questions RQ1 Is it possible to define a model, based on software engineering techniques, for representing the personalised health domain in an abstract and representative way? RQ2 Is it possible to build a development platform based on this model? RQ3 Does the development platform help developers create complex integrated pHealth systems? Methods As method for describing the model, the ISO/IEC/IEEE 42010 framework [25] is adopted for its generality and high level of abstraction. The model is specified in different parts: a conceptual model, which makes use of concept maps, for representing stakeholders, artefacts and shared information, and in scenarios and use cases for the representation of the functionalities of pHealth systems. The model was derived from literature analysis, including 7 industrial and scientific reports, 9 electronic standards, 10 conference proceedings papers, 37 journal papers, 25 websites and 5 books. Based on the reference model, requirements were drawn for building the development platform enriched with a set of requirements gathered in a survey run among 11 experienced engineers. For developing the platform, the continuous integration methodology [74] was adopted which allowed to perform automatic tests on a server and also to deploy packaged releases on a web site. As a validation methodology, a theory building framework for SW engineering was adopted from [181]. The framework, chosen as a guide to find evidence for justifying the research questions, imposed the creation of theories based on models and propositions to be validated within a scope. The validation of the model was conducted as an on-line survey in three validation rounds, encompassing a growing number of participants. The survey was submitted to 134 experts of the field and on some public channels like relevant mailing lists and social networks. Its objective was to assess the model's readability, its level of coverage of the domain and its potential usefulness in the design of actual, derived systems. The questionnaires included quantitative Likert scale questions and free text inputs for comments. The development platform was validated in two scopes. As a small-scale experiment, the platform was used in a 12 hours training session where 4 developers had to perform an exercise consisting in developing a set of typical pHealth use cases At the end of the session, a focus group was held to identify benefits and drawbacks of the platform. The second validation was held as a test-case study in a large scale research project called HeartCycle the aim of which was to develop a closed-loop disease management system for heart failure and coronary heart disease patients [160]. During this project three applications were developed by a team of programmers and designers. One of these applications was tested in a clinical trial with actual patients. At the end of the project, the team was interviewed in a focus group to assess the role the platform had within the project. Results For what regards the model that describes the pHealth domain, its conceptual part includes a description of the main roles and concerns of pHealth stakeholders, a model of the ICT artefacts that are commonly adopted and a model representing the typical data that need to be formalized among pHealth systems. The functional model includes a set of 18 scenarios, divided into assisted person's view, caregiver's view, developer's view, technology and services providers' view and authority's view, and a set of 52 Use Cases grouped in 6 categories: assisted person's activities, system reactions, caregiver's activities, user engagement, developer's activities and deployer's activities. For what concerns the validation of the model, a total of 65 people participated in the online survey providing their level of agreement in all the assessed dimensions and a total of 248 comments on how to improve and complete the model. Participants' background spanned from engineering and software development (70%) to medical specialities (15%), with declared interest in the fields of eHealth (24%), mHealth (16%), Ambient Assisted Living (21%), Personalized Medicine (5%), Personal Health Systems (15%), Medical Informatics (10%) and Biomedical Engineering (8%) with an average of 7.25_4.99 years of experience in these fields. From the analysis of the answers it is possible to observe that the contacted experts considered the model easily readable (average of 1.89_0.79 being 1 the most favourable scoring and 5 the worst), sufficiently abstract (1.99_0.88) and formal (2.13_0.77) for its purpose, with a sufficient coverage of the domain (2.26_0.95), useful for describing the domain (2.02_0.7) and for generating more specific systems (2_0.75) and they reported a partial interest in using the model in their job (2.48_0.91). Thanks to their comments, the model was improved and enriched with concepts that were missing at the beginning, nonetheless it was not possible to prove an improvement among the iterations, due to the diversity of the participants in the three rounds. From the model, a development platform for the pHealth domain was generated called pHealth Patient Platform (pHPP). The platform includes a set of libraries, programming and deployment tools, a tutorial and a sample application. The main four modules of the architecture are: the Data Collection Engine, which allows abstracting sources of information like sensors or external services, mapping data to databases and ontologies, and allowing event-based interaction and filtering, the GUI Engine, which abstracts the user interface in a message-like interaction model, the Workow Engine, which allows programming the application's user interaction ows with graphical workows, and the Rule Engine, which gives developers a simple means for programming the application's logic in the form of \if-then" rules. After the 5 years experience of HeartCycle, partially programmed with pHPP, 5 developers were joined in a focus group to discuss the advantages and drawbacks of the platform. The view that emerged from the training course and the focus group was that the platform is well-suited to the needs of the engineers working in the field, it allowed the separation of concerns among the different specialities and it simplified some common development tasks like data management and asynchronous interaction. Nevertheless, some deficiencies were pointed out in terms of a lack of maturity of some technological choices, and for the absence of some domain-specific tools, e.g. for data processing or for health-related communication protocols. Within HeartCycle, the platform was used to develop part of the Guided Exercise system, a composition of ICT tools for the physical rehabilitation of patients who suffered from myocardial infarction. The system developed using the platform was tested in a randomized controlled clinical trial, in which 55 patients used the system for 21 weeks. The technical results of this trial showed that the system was stable and reliable. Some minor bugs were detected, but these were promptly corrected using the platform. This shows that the platform, as well as facilitating the development task, can be successfully used to produce reliable software. Conclusions The research work carried out in developing this thesis provides responses to the three three research questions that were the motivation for the work. RQ1 A model was developed representing the domain of personalised health systems, and the assessment of experts in the field was that it represents the domain accurately, with an appropriate balance between abstraction and detail. RQ2 A development platform based on the model was successfully developed. RQ3 The platform has been shown to assist developers create complex pHealth software. This was demonstrated within the scope of one large-scale project, but the generic approach adopted provides indications that it would offer benefits more widely. The results of these evaluations provide indications that both the model and the platform are good candidates for being a reference for future pHealth developments.

BioMet®Phon: A system to monitor phonation quality in the clinics

BioMet®Phon is a software application developed for the characterization of voice in voice quality evaluation. Initially it was conceived as plain research code to estimate the glottal source from voice and obtain the biomechanical parameters of the vocal folds from the spectral density of the estimate. This code grew to what is now the Glottex®Engine package (G®E). Further demands from users in laryngology and speech therapy fields instantiated the development of a specific Graphic User Interface (GUI’s) to encapsulate user interaction with the G®E. This gave place to BioMet®Phon, an application which extracts the glottal source from voice and offers a complete parameterization of this signal, including distortion, cepstral, spectral, biomechanical, time domain, contact and tremor parameters. The semantic capabilities of biomechanical parameters are discussed. Study cases from its application to the field of laryngology and speech therapy are given and discussed. Validation results in voice pathology detection are also presented. Applications to laryngology, speech therapy, and monitoring neurological deterioration in the elder are proposed.