Automated detection and classification of brain injury lesions on structural magnetic resonance imaging

El daño cerebral adquirido (DCA) es un problema social y sanitario grave, de magnitud creciente y de una gran complejidad diagnóstica y terapéutica. Su elevada incidencia, junto con el aumento de la supervivencia de los pacientes, una vez superada la fase aguda, lo convierten también en un problema de alta prevalencia. En concreto, según la Organización Mundial de la Salud (OMS) el DCA estará entre las 10 causas más comunes de discapacidad en el año 2020. La neurorrehabilitación permite mejorar el déficit tanto cognitivo como funcional y aumentar la autonomía de las personas con DCA. Con la incorporación de nuevas soluciones tecnológicas al proceso de neurorrehabilitación se pretende alcanzar un nuevo paradigma donde se puedan diseñar tratamientos que sean intensivos, personalizados, monitorizados y basados en la evidencia. Ya que son estas cuatro características las que aseguran que los tratamientos son eficaces. A diferencia de la mayor parte de las disciplinas médicas, no existen asociaciones de síntomas y signos de la alteración cognitiva que faciliten la orientación terapéutica. Actualmente, los tratamientos de neurorrehabilitación se diseñan en base a los resultados obtenidos en una batería de evaluación neuropsicológica que evalúa el nivel de afectación de cada una de las funciones cognitivas (memoria, atención, funciones ejecutivas, etc.). La línea de investigación en la que se enmarca este trabajo de investigación pretende diseñar y desarrollar un perfil cognitivo basado no sólo en el resultado obtenido en esa batería de test, sino también en información teórica que engloba tanto estructuras anatómicas como relaciones funcionales e información anatómica obtenida de los estudios de imagen. De esta forma, el perfil cognitivo utilizado para diseñar los tratamientos integra información personalizada y basada en la evidencia. Las técnicas de neuroimagen representan una herramienta fundamental en la identificación de lesiones para la generación de estos perfiles cognitivos. La aproximación clásica utilizada en la identificación de lesiones consiste en delinear manualmente regiones anatómicas cerebrales. Esta aproximación presenta diversos problemas relacionados con inconsistencias de criterio entre distintos clínicos, reproducibilidad y tiempo. Por tanto, la automatización de este procedimiento es fundamental para asegurar una extracción objetiva de información. La delineación automática de regiones anatómicas se realiza mediante el registro tanto contra atlas como contra otros estudios de imagen de distintos sujetos. Sin embargo, los cambios patológicos asociados al DCA están siempre asociados a anormalidades de intensidad y/o cambios en la localización de las estructuras. Este hecho provoca que los algoritmos de registro tradicionales basados en intensidad no funcionen correctamente y requieran la intervención del clínico para seleccionar ciertos puntos (que en esta tesis hemos denominado puntos singulares). Además estos algoritmos tampoco permiten que se produzcan deformaciones grandes deslocalizadas. Hecho que también puede ocurrir ante la presencia de lesiones provocadas por un accidente cerebrovascular (ACV) o un traumatismo craneoencefálico (TCE). Esta tesis se centra en el diseño, desarrollo e implementación de una metodología para la detección automática de estructuras lesionadas que integra algoritmos cuyo objetivo principal es generar resultados que puedan ser reproducibles y objetivos. Esta metodología se divide en cuatro etapas: pre-procesado, identificación de puntos singulares, registro y detección de lesiones. Los trabajos y resultados alcanzados en esta tesis son los siguientes: Pre-procesado. En esta primera etapa el objetivo es homogeneizar todos los datos de entrada con el objetivo de poder extraer conclusiones válidas de los resultados obtenidos. Esta etapa, por tanto, tiene un gran impacto en los resultados finales. Se compone de tres operaciones: eliminación del cráneo, normalización en intensidad y normalización espacial. Identificación de puntos singulares. El objetivo de esta etapa es automatizar la identificación de puntos anatómicos (puntos singulares). Esta etapa equivale a la identificación manual de puntos anatómicos por parte del clínico, permitiendo: identificar un mayor número de puntos lo que se traduce en mayor información; eliminar el factor asociado a la variabilidad inter-sujeto, por tanto, los resultados son reproducibles y objetivos; y elimina el tiempo invertido en el marcado manual de puntos. Este trabajo de investigación propone un algoritmo de identificación de puntos singulares (descriptor) basado en una solución multi-detector y que contiene información multi-paramétrica: espacial y asociada a la intensidad. Este algoritmo ha sido contrastado con otros algoritmos similares encontrados en el estado del arte. Registro. En esta etapa se pretenden poner en concordancia espacial dos estudios de imagen de sujetos/pacientes distintos. El algoritmo propuesto en este trabajo de investigación está basado en descriptores y su principal objetivo es el cálculo de un campo vectorial que permita introducir deformaciones deslocalizadas en la imagen (en distintas regiones de la imagen) y tan grandes como indique el vector de deformación asociado. El algoritmo propuesto ha sido comparado con otros algoritmos de registro utilizados en aplicaciones de neuroimagen que se utilizan con estudios de sujetos control. Los resultados obtenidos son prometedores y representan un nuevo contexto para la identificación automática de estructuras. Identificación de lesiones. En esta última etapa se identifican aquellas estructuras cuyas características asociadas a la localización espacial y al área o volumen han sido modificadas con respecto a una situación de normalidad. Para ello se realiza un estudio estadístico del atlas que se vaya a utilizar y se establecen los parámetros estadísticos de normalidad asociados a la localización y al área. En función de las estructuras delineadas en el atlas, se podrán identificar más o menos estructuras anatómicas, siendo nuestra metodología independiente del atlas seleccionado. En general, esta tesis doctoral corrobora las hipótesis de investigación postuladas relativas a la identificación automática de lesiones utilizando estudios de imagen médica estructural, concretamente estudios de resonancia magnética. Basándose en estos cimientos, se han abrir nuevos campos de investigación que contribuyan a la mejora en la detección de lesiones. ABSTRACT Brain injury constitutes a serious social and health problem of increasing magnitude and of great diagnostic and therapeutic complexity. Its high incidence and survival rate, after the initial critical phases, makes it a prevalent problem that needs to be addressed. In particular, according to the World Health Organization (WHO), brain injury will be among the 10 most common causes of disability by 2020. Neurorehabilitation improves both cognitive and functional deficits and increases the autonomy of brain injury patients. The incorporation of new technologies to the neurorehabilitation tries to reach a new paradigm focused on designing intensive, personalized, monitored and evidence-based treatments. Since these four characteristics ensure the effectivity of treatments. Contrary to most medical disciplines, it is not possible to link symptoms and cognitive disorder syndromes, to assist the therapist. Currently, neurorehabilitation treatments are planned considering the results obtained from a neuropsychological assessment battery, which evaluates the functional impairment of each cognitive function (memory, attention, executive functions, etc.). The research line, on which this PhD falls under, aims to design and develop a cognitive profile based not only on the results obtained in the assessment battery, but also on theoretical information that includes both anatomical structures and functional relationships and anatomical information obtained from medical imaging studies, such as magnetic resonance. Therefore, the cognitive profile used to design these treatments integrates information personalized and evidence-based. Neuroimaging techniques represent an essential tool to identify lesions and generate this type of cognitive dysfunctional profiles. Manual delineation of brain anatomical regions is the classical approach to identify brain anatomical regions. Manual approaches present several problems related to inconsistencies across different clinicians, time and repeatability. Automated delineation is done by registering brains to one another or to a template. However, when imaging studies contain lesions, there are several intensity abnormalities and location alterations that reduce the performance of most of the registration algorithms based on intensity parameters. Thus, specialists may have to manually interact with imaging studies to select landmarks (called singular points in this PhD) or identify regions of interest. These two solutions have the same inconvenient than manual approaches, mentioned before. Moreover, these registration algorithms do not allow large and distributed deformations. This type of deformations may also appear when a stroke or a traumatic brain injury (TBI) occur. This PhD is focused on the design, development and implementation of a new methodology to automatically identify lesions in anatomical structures. This methodology integrates algorithms whose main objective is to generate objective and reproducible results. It is divided into four stages: pre-processing, singular points identification, registration and lesion detection. Pre-processing stage. In this first stage, the aim is to standardize all input data in order to be able to draw valid conclusions from the results. Therefore, this stage has a direct impact on the final results. It consists of three steps: skull-stripping, spatial and intensity normalization. Singular points identification. This stage aims to automatize the identification of anatomical points (singular points). It involves the manual identification of anatomical points by the clinician. This automatic identification allows to identify a greater number of points which results in more information; to remove the factor associated to inter-subject variability and thus, the results are reproducible and objective; and to eliminate the time spent on manual marking. This PhD proposed an algorithm to automatically identify singular points (descriptor) based on a multi-detector approach. This algorithm contains multi-parametric (spatial and intensity) information. This algorithm has been compared with other similar algorithms found on the state of the art. Registration. The goal of this stage is to put in spatial correspondence two imaging studies of different subjects/patients. The algorithm proposed in this PhD is based on descriptors. Its main objective is to compute a vector field to introduce distributed deformations (changes in different imaging regions), as large as the deformation vector indicates. The proposed algorithm has been compared with other registration algorithms used on different neuroimaging applications which are used with control subjects. The obtained results are promising and they represent a new context for the automatic identification of anatomical structures. Lesion identification. This final stage aims to identify those anatomical structures whose characteristics associated to spatial location and area or volume has been modified with respect to a normal state. A statistical study of the atlas to be used is performed to establish which are the statistical parameters associated to the normal state. The anatomical structures that may be identified depend on the selected anatomical structures identified on the atlas. The proposed methodology is independent from the selected atlas. Overall, this PhD corroborates the investigated research hypotheses regarding the automatic identification of lesions based on structural medical imaging studies (resonance magnetic studies). Based on these foundations, new research fields to improve the automatic identification of lesions in brain injury can be proposed.

Proposal of a trust and reputation model to provide recommendation in open quality oriented environments

El auge y penetración de las nuevas tecnologías junto con la llamada Web Social están cambiando la forma en la que accedemos a la medicina. Cada vez más pacientes y profesionales de la medicina están creando y consumiendo recursos digitales de contenido clínico a través de Internet, surgiendo el problema de cómo asegurar la fiabilidad de estos recursos. Además, un nuevo concepto está apareciendo, el de pervasive healthcare o sanidad ubicua, motivado por pacientes que demandan un acceso a los servicios sanitarios en todo momento y en todo lugar. Este nuevo escenario lleva aparejado un problema de confianza en los proveedores de servicios sanitarios. Las plataformas de eLearning se están erigiendo como paradigma de esta nueva Medicina 2.0 ya que proveen un servicio abierto a la vez que controlado/supervisado a recursos digitales, y facilitan las interacciones y consultas entre usuarios, suponiendo una buena aproximación para esta sanidad ubicua. En estos entornos los problemas de fiabilidad y confianza pueden ser solventados mediante la implementación de mecanismos de recomendación de recursos y personas de manera confiable. Tradicionalmente las plataformas de eLearning ya cuentan con mecanismos de recomendación, si bien están más enfocados a la recomendación de recursos. Para la recomendación de usuarios es necesario acudir a mecanismos más elaborados como son los sistemas de confianza y reputación (trust and reputation) En ambos casos, tanto la recomendación de recursos como el cálculo de la reputación de los usuarios se realiza teniendo en cuenta criterios principalmente subjetivos como son las opiniones de los usuarios. En esta tesis doctoral proponemos un nuevo modelo de confianza y reputación que combina evaluaciones automáticas de los recursos digitales en una plataforma de eLearning, con las opiniones vertidas por los usuarios como resultado de las interacciones con otros usuarios o después de consumir un recurso. El enfoque seguido presenta la novedad de la combinación de una parte objetiva con otra subjetiva, persiguiendo mitigar el efecto de posibles castigos subjetivos por parte de usuarios malintencionados, a la vez que enriquecer las evaluaciones objetivas con información adicional acerca de la capacidad pedagógica del recurso o de la persona. El resultado son recomendaciones siempre adaptadas a los requisitos de los usuarios, y de la máxima calidad tanto técnica como educativa. Esta nueva aproximación requiere una nueva herramienta para su validación in-silico, al no existir ninguna aplicación que permita la simulación de plataformas de eLearning con mecanismos de recomendación de recursos y personas, donde además los recursos sean evaluados objetivamente. Este trabajo de investigación propone pues una nueva herramienta, basada en el paradigma de programación orientada a agentes inteligentes para el modelado de comportamientos complejos de usuarios en plataformas de eLearning. Además, la herramienta permite también la simulación del funcionamiento de este tipo de entornos dedicados al intercambio de conocimiento. La evaluación del trabajo propuesto en este documento de tesis se ha realizado de manera iterativa a lo largo de diferentes escenarios en los que se ha situado al sistema frente a una amplia gama de comportamientos de usuarios. Se ha comparado el rendimiento del modelo de confianza y reputación propuesto frente a dos modos de recomendación tradicionales: a) utilizando sólo las opiniones subjetivas de los usuarios para el cálculo de la reputación y por extensión la recomendación; y b) teniendo en cuenta sólo la calidad objetiva del recurso sin hacer ningún cálculo de reputación. Los resultados obtenidos nos permiten afirmar que el modelo desarrollado mejora la recomendación ofrecida por las aproximaciones tradicionales, mostrando una mayor flexibilidad y capacidad de adaptación a diferentes situaciones. Además, el modelo propuesto es capaz de asegurar la recomendación de nuevos usuarios entrando al sistema frente a la nula recomendación para estos usuarios presentada por el modo de recomendación predominante en otras plataformas que basan la recomendación sólo en las opiniones de otros usuarios. Por último, el paradigma de agentes inteligentes ha probado su valía a la hora de modelar plataformas virtuales complejas orientadas al intercambio de conocimiento, especialmente a la hora de modelar y simular el comportamiento de los usuarios de estos entornos. La herramienta de simulación desarrollada ha permitido la evaluación del modelo de confianza y reputación propuesto en esta tesis en una amplia gama de situaciones diferentes. ABSTRACT Internet is changing everything, and this revolution is especially present in traditionally offline spaces such as medicine. In recent years health consumers and health service providers are actively creating and consuming Web contents stimulated by the emergence of the Social Web. Reliability stands out as the main concern when accessing the overwhelming amount of information available online. Along with this new way of accessing the medicine, new concepts like ubiquitous or pervasive healthcare are appearing. Trustworthiness assessment is gaining relevance: open health provisioning systems require mechanisms that help evaluating individuals’ reputation in pursuit of introducing safety to these open and dynamic environments. Technical Enhanced Learning (TEL) -commonly known as eLearning- platforms arise as a paradigm of this Medicine 2.0. They provide an open while controlled/supervised access to resources generated and shared by users, enhancing what it is being called informal learning. TEL systems also facilitate direct interactions amongst users for consultation, resulting in a good approach to ubiquitous healthcare. The aforementioned reliability and trustworthiness problems can be faced by the implementation of mechanisms for the trusted recommendation of both resources and healthcare services providers. Traditionally, eLearning platforms already integrate recommendation mechanisms, although this recommendations are basically focused on providing an ordered classifications of resources. For users’ recommendation, the implementation of trust and reputation systems appears as the best solution. Nevertheless, both approaches base the recommendation on the information from the subjective opinions of other users of the platform regarding the resources or the users. In this PhD work a novel approach is presented for the recommendation of both resources and users within open environments focused on knowledge exchange, as it is the case of TEL systems for ubiquitous healthcare. The proposed solution adds the objective evaluation of the resources to the traditional subjective personal opinions to estimate the reputation of the resources and of the users of the system. This combined measure, along with the reliability of that calculation, is used to provide trusted recommendations. The integration of opinions and evaluations, subjective and objective, allows the model to defend itself against misbehaviours. Furthermore, it also allows ‘colouring’ cold evaluation values by providing additional quality information such as the educational capacities of a digital resource in an eLearning system. As a result, the recommendations are always adapted to user requirements, and of the maximum technical and educational quality. To our knowledge, the combination of objective assessments and subjective opinions to provide recommendation has not been considered before in the literature. Therefore, for the evaluation of the trust and reputation model defined in this PhD thesis, a new simulation tool will be developed following the agent-oriented programming paradigm. The multi-agent approach allows an easy modelling of independent and proactive behaviours for the simulation of users of the system, conforming a faithful resemblance of real users of TEL platforms. For the evaluation of the proposed work, an iterative approach have been followed, testing the performance of the trust and reputation model while providing recommendation in a varied range of scenarios. A comparison with two traditional recommendation mechanisms was performed: a) using only users’ past opinions about a resource and/or other users; and b) not using any reputation assessment and providing the recommendation considering directly the objective quality of the resources. The results show that the developed model improves traditional approaches at providing recommendations in Technology Enhanced Learning (TEL) platforms, presenting a higher adaptability to different situations, whereas traditional approaches only have good results under favourable conditions. Furthermore the promotion period mechanism implemented successfully helps new users in the system to be recommended for direct interactions as well as the resources created by them. On the contrary OnlyOpinions fails completely and new users are never recommended, while traditional approaches only work partially. Finally, the agent-oriented programming (AOP) paradigm has proven its validity at modelling users’ behaviours in TEL platforms. Intelligent software agents’ characteristics matched the main requirements of the simulation tool. The proactivity, sociability and adaptability of the developed agents allowed reproducing real users’ actions and attitudes through the diverse situations defined in the evaluation framework. The result were independent users, accessing to different resources and communicating amongst them to fulfil their needs, basing these interactions on the recommendations provided by the reputation engine.

Trust-ware: A Methodology to Analyze, Design, and Secure Trust and Reputation Systems

Entendemos por inteligencia colectiva una forma de inteligencia que surge de la colaboración y la participación de varios individuos o, siendo más estrictos, varias entidades. En base a esta sencilla definición podemos observar que este concepto es campo de estudio de las más diversas disciplinas como pueden ser la sociología, las tecnologías de la información o la biología, atendiendo cada una de ellas a un tipo de entidades diferentes: seres humanos, elementos de computación o animales. Como elemento común podríamos indicar que la inteligencia colectiva ha tenido como objetivo el ser capaz de fomentar una inteligencia de grupo que supere a la inteligencia individual de las entidades que lo forman a través de mecanismos de coordinación, cooperación, competencia, integración, diferenciación, etc. Sin embargo, aunque históricamente la inteligencia colectiva se ha podido desarrollar de forma paralela e independiente en las distintas disciplinas que la tratan, en la actualidad, los avances en las tecnologías de la información han provocado que esto ya no sea suficiente. Hoy en día seres humanos y máquinas a través de todo tipo de redes de comunicación e interfaces, conviven en un entorno en el que la inteligencia colectiva ha cobrado una nueva dimensión: ya no sólo puede intentar obtener un comportamiento superior al de sus entidades constituyentes sino que ahora, además, estas inteligencias individuales son completamente diferentes unas de otras y aparece por lo tanto el doble reto de ser capaces de gestionar esta gran heterogeneidad y al mismo tiempo ser capaces de obtener comportamientos aún más inteligentes gracias a las sinergias que los distintos tipos de inteligencias pueden generar. Dentro de las áreas de trabajo de la inteligencia colectiva existen varios campos abiertos en los que siempre se intenta obtener unas prestaciones superiores a las de los individuos. Por ejemplo: consciencia colectiva, memoria colectiva o sabiduría colectiva. Entre todos estos campos nosotros nos centraremos en uno que tiene presencia en la práctica totalidad de posibles comportamientos inteligentes: la toma de decisiones. El campo de estudio de la toma de decisiones es realmente amplio y dentro del mismo la evolución ha sido completamente paralela a la que citábamos anteriormente en referencia a la inteligencia colectiva. En primer lugar se centró en el individuo como entidad decisoria para posteriormente desarrollarse desde un punto de vista social, institucional, etc. La primera fase dentro del estudio de la toma de decisiones se basó en la utilización de paradigmas muy sencillos: análisis de ventajas e inconvenientes, priorización basada en la maximización de algún parámetro del resultado, capacidad para satisfacer los requisitos de forma mínima por parte de las alternativas, consultas a expertos o entidades autorizadas o incluso el azar. Sin embargo, al igual que el paso del estudio del individuo al grupo supone una nueva dimensión dentro la inteligencia colectiva la toma de decisiones colectiva supone un nuevo reto en todas las disciplinas relacionadas. Además, dentro de la decisión colectiva aparecen dos nuevos frentes: los sistemas de decisión centralizados y descentralizados. En el presente proyecto de tesis nos centraremos en este segundo, que es el que supone una mayor atractivo tanto por las posibilidades de generar nuevo conocimiento y trabajar con problemas abiertos actualmente así como en lo que respecta a la aplicabilidad de los resultados que puedan obtenerse. Ya por último, dentro del campo de los sistemas de decisión descentralizados existen varios mecanismos fundamentales que dan lugar a distintas aproximaciones a la problemática propia de este campo. Por ejemplo el liderazgo, la imitación, la prescripción o el miedo. Nosotros nos centraremos en uno de los más multidisciplinares y con mayor capacidad de aplicación en todo tipo de disciplinas y que, históricamente, ha demostrado que puede dar lugar a prestaciones muy superiores a otros tipos de mecanismos de decisión descentralizados: la confianza y la reputación. Resumidamente podríamos indicar que confianza es la creencia por parte de una entidad que otra va a realizar una determinada actividad de una forma concreta. En principio es algo subjetivo, ya que la confianza de dos entidades diferentes sobre una tercera no tiene porqué ser la misma. Por otro lado, la reputación es la idea colectiva (o evaluación social) que distintas entidades de un sistema tiene sobre otra entidad del mismo en lo que respecta a un determinado criterio. Es por tanto una información de carácter colectivo pero única dentro de un sistema, no asociada a cada una de las entidades del sistema sino por igual a todas ellas. En estas dos sencillas definiciones se basan la inmensa mayoría de sistemas colectivos. De hecho muchas disertaciones indican que ningún tipo de organización podría ser viable de no ser por la existencia y la utilización de los conceptos de confianza y reputación. A partir de ahora, a todo sistema que utilice de una u otra forma estos conceptos lo denominaremos como sistema de confianza y reputación (o TRS, Trust and Reputation System). Sin embargo, aunque los TRS son uno de los aspectos de nuestras vidas más cotidianos y con un mayor campo de aplicación, el conocimiento que existe actualmente sobre ellos no podría ser más disperso. Existen un gran número de trabajos científicos en todo tipo de áreas de conocimiento: filosofía, psicología, sociología, economía, política, tecnologías de la información, etc. Pero el principal problema es que no existe una visión completa de la confianza y reputación en su sentido más amplio. Cada disciplina focaliza sus estudios en unos aspectos u otros dentro de los TRS, pero ninguna de ellas trata de explotar el conocimiento generado en el resto para mejorar sus prestaciones en su campo de aplicación concreto. Aspectos muy detallados en algunas áreas de conocimiento son completamente obviados por otras, o incluso aspectos tratados por distintas disciplinas, al ser estudiados desde distintos puntos de vista arrojan resultados complementarios que, sin embargo, no son aprovechados fuera de dichas áreas de conocimiento. Esto nos lleva a una dispersión de conocimiento muy elevada y a una falta de reutilización de metodologías, políticas de actuación y técnicas de una disciplina a otra. Debido su vital importancia, esta alta dispersión de conocimiento se trata de uno de los principales problemas que se pretenden resolver con el presente trabajo de tesis. Por otro lado, cuando se trabaja con TRS, todos los aspectos relacionados con la seguridad están muy presentes ya que muy este es un tema vital dentro del campo de la toma de decisiones. Además también es habitual que los TRS se utilicen para desempeñar responsabilidades que aportan algún tipo de funcionalidad relacionada con el mundo de la seguridad. Por último no podemos olvidar que el acto de confiar está indefectiblemente unido al de delegar una determinada responsabilidad, y que al tratar estos conceptos siempre aparece la idea de riesgo, riesgo de que las expectativas generadas por el acto de la delegación no se cumplan o se cumplan de forma diferente. Podemos ver por lo tanto que cualquier sistema que utiliza la confianza para mejorar o posibilitar su funcionamiento, por su propia naturaleza, es especialmente vulnerable si las premisas en las que se basa son atacadas. En este sentido podemos comprobar (tal y como analizaremos en más detalle a lo largo del presente documento) que las aproximaciones que realizan las distintas disciplinas que tratan la violación de los sistemas de confianza es de lo más variado. únicamente dentro del área de las tecnologías de la información se ha intentado utilizar alguno de los enfoques de otras disciplinas de cara a afrontar problemas relacionados con la seguridad de TRS. Sin embargo se trata de una aproximación incompleta y, normalmente, realizada para cumplir requisitos de aplicaciones concretas y no con la idea de afianzar una base de conocimiento más general y reutilizable en otros entornos. Con todo esto en cuenta, podemos resumir contribuciones del presente trabajo de tesis en las siguientes. • La realización de un completo análisis del estado del arte dentro del mundo de la confianza y la reputación que nos permite comparar las ventajas e inconvenientes de las diferentes aproximación que se realizan a estos conceptos en distintas áreas de conocimiento. • La definición de una arquitectura de referencia para TRS que contempla todas las entidades y procesos que intervienen en este tipo de sistemas. • La definición de un marco de referencia para analizar la seguridad de TRS. Esto implica tanto identificar los principales activos de un TRS en lo que respecta a la seguridad, así como el crear una tipología de posibles ataques y contramedidas en base a dichos activos. • La propuesta de una metodología para el análisis, el diseño, el aseguramiento y el despliegue de un TRS en entornos reales. Adicionalmente se exponen los principales tipos de aplicaciones que pueden obtenerse de los TRS y los medios para maximizar sus prestaciones en cada una de ellas. • La generación de un software que permite simular cualquier tipo de TRS en base a la arquitectura propuesta previamente. Esto permite evaluar las prestaciones de un TRS bajo una determinada configuración en un entorno controlado previamente a su despliegue en un entorno real. Igualmente es de gran utilidad para evaluar la resistencia a distintos tipos de ataques o mal-funcionamientos del sistema. Además de las contribuciones realizadas directamente en el campo de los TRS, hemos realizado aportaciones originales a distintas áreas de conocimiento gracias a la aplicación de las metodologías de análisis y diseño citadas con anterioridad. • Detección de anomalías térmicas en Data Centers. Hemos implementado con éxito un sistema de deteción de anomalías térmicas basado en un TRS. Comparamos la detección de prestaciones de algoritmos de tipo Self-Organized Maps (SOM) y Growing Neural Gas (GNG). Mostramos como SOM ofrece mejores resultados para anomalías en los sistemas de refrigeración de la sala mientras que GNG es una opción más adecuada debido a sus tasas de detección y aislamiento para casos de anomalías provocadas por una carga de trabajo excesiva. • Mejora de las prestaciones de recolección de un sistema basado en swarm computing y odometría social. Gracias a la implementación de un TRS conseguimos mejorar las capacidades de coordinación de una red de robots autónomos distribuidos. La principal contribución reside en el análisis y la validación de las mejoras increméntales que pueden conseguirse con la utilización apropiada de la información existente en el sistema y que puede ser relevante desde el punto de vista de un TRS, y con la implementación de algoritmos de cálculo de confianza basados en dicha información. • Mejora de la seguridad de Wireless Mesh Networks contra ataques contra la integridad, la confidencialidad o la disponibilidad de los datos y / o comunicaciones soportadas por dichas redes. • Mejora de la seguridad de Wireless Sensor Networks contra ataques avanzamos, como insider attacks, ataques desconocidos, etc. Gracias a las metodologías presentadas implementamos contramedidas contra este tipo de ataques en entornos complejos. En base a los experimentos realizados, hemos demostrado que nuestra aproximación es capaz de detectar y confinar varios tipos de ataques que afectan a los protocoles esenciales de la red. La propuesta ofrece unas velocidades de detección muy altas así como demuestra que la inclusión de estos mecanismos de actuación temprana incrementa significativamente el esfuerzo que un atacante tiene que introducir para comprometer la red. Finalmente podríamos concluir que el presente trabajo de tesis supone la generación de un conocimiento útil y aplicable a entornos reales, que nos permite la maximización de las prestaciones resultantes de la utilización de TRS en cualquier tipo de campo de aplicación. De esta forma cubrimos la principal carencia existente actualmente en este campo, que es la falta de una base de conocimiento común y agregada y la inexistencia de una metodología para el desarrollo de TRS que nos permita analizar, diseñar, asegurar y desplegar TRS de una forma sistemática y no artesanal y ad-hoc como se hace en la actualidad. ABSTRACT By collective intelligence we understand a form of intelligence that emerges from the collaboration and competition of many individuals, or strictly speaking, many entities. Based on this simple definition, we can see how this concept is the field of study of a wide range of disciplines, such as sociology, information science or biology, each of them focused in different kinds of entities: human beings, computational resources, or animals. As a common factor, we can point that collective intelligence has always had the goal of being able of promoting a group intelligence that overcomes the individual intelligence of the basic entities that constitute it. This can be accomplished through different mechanisms such as coordination, cooperation, competence, integration, differentiation, etc. Collective intelligence has historically been developed in a parallel and independent way among the different disciplines that deal with it. However, this is not enough anymore due to the advances in information technologies. Nowadays, human beings and machines coexist in environments where collective intelligence has taken a new dimension: we yet have to achieve a better collective behavior than the individual one, but now we also have to deal with completely different kinds of individual intelligences. Therefore, we have a double goal: being able to deal with this heterogeneity and being able to get even more intelligent behaviors thanks to the synergies that the different kinds of intelligence can generate. Within the areas of collective intelligence there are several open topics where they always try to get better performances from groups than from the individuals. For example: collective consciousness, collective memory, or collective wisdom. Among all these topics we will focus on collective decision making, that has influence in most of the collective intelligent behaviors. The field of study of decision making is really wide, and its evolution has been completely parallel to the aforementioned collective intelligence. Firstly, it was focused on the individual as the main decision-making entity, but later it became involved in studying social and institutional groups as basic decision-making entities. The first studies within the decision-making discipline were based on simple paradigms, such as pros and cons analysis, criteria prioritization, fulfillment, following orders, or even chance. However, in the same way that studying the community instead of the individual meant a paradigm shift within collective intelligence, collective decision-making means a new challenge for all the related disciplines. Besides, two new main topics come up when dealing with collective decision-making: centralized and decentralized decision-making systems. In this thesis project we focus in the second one, because it is the most interesting based on the opportunities to generate new knowledge and deal with open issues in this area, as well as these results can be put into practice in a wider set of real-life environments. Finally, within the decentralized collective decision-making systems discipline, there are several basic mechanisms that lead to different approaches to the specific problems of this field, for example: leadership, imitation, prescription, or fear. We will focus on trust and reputation. They are one of the most multidisciplinary concepts and with more potential for applying them in every kind of environments. Besides, they have historically shown that they can generate better performance than other decentralized decision-making mechanisms. Shortly, we say trust is the belief of one entity that the outcome of other entities’ actions is going to be in a specific way. It is a subjective concept because the trust of two different entities in another one does not have to be the same. Reputation is the collective idea (or social evaluation) that a group of entities within a system have about another entity based on a specific criterion. Thus, it is a collective concept in its origin. It is important to say that the behavior of most of the collective systems are based on these two simple definitions. In fact, a lot of articles and essays describe how any organization would not be viable if the ideas of trust and reputation did not exist. From now on, we call Trust an Reputation System (TRS) to any kind of system that uses these concepts. Even though TRSs are one of the most common everyday aspects in our lives, the existing knowledge about them could not be more dispersed. There are thousands of scientific works in every field of study related to trust and reputation: philosophy, psychology, sociology, economics, politics, information sciences, etc. But the main issue is that a comprehensive vision of trust and reputation for all these disciplines does not exist. Every discipline focuses its studies on a specific set of topics but none of them tries to take advantage of the knowledge generated in the other disciplines to improve its behavior or performance. Detailed topics in some fields are completely obviated in others, and even though the study of some topics within several disciplines produces complementary results, these results are not used outside the discipline where they were generated. This leads us to a very high knowledge dispersion and to a lack in the reuse of methodologies, policies and techniques among disciplines. Due to its great importance, this high dispersion of trust and reputation knowledge is one of the main problems this thesis contributes to solve. When we work with TRSs, all the aspects related to security are a constant since it is a vital aspect within the decision-making systems. Besides, TRS are often used to perform some responsibilities related to security. Finally, we cannot forget that the act of trusting is invariably attached to the act of delegating a specific responsibility and, when we deal with these concepts, the idea of risk is always present. This refers to the risk of generated expectations not being accomplished or being accomplished in a different way we anticipated. Thus, we can see that any system using trust to improve or enable its behavior, because of its own nature, is especially vulnerable if the premises it is based on are attacked. Related to this topic, we can see that the approaches of the different disciplines that study attacks of trust and reputation are very diverse. Some attempts of using approaches of other disciplines have been made within the information science area of knowledge, but these approaches are usually incomplete, not systematic and oriented to achieve specific requirements of specific applications. They never try to consolidate a common base of knowledge that could be reusable in other context. Based on all these ideas, this work makes the following direct contributions to the field of TRS: • The compilation of the most relevant existing knowledge related to trust and reputation management systems focusing on their advantages and disadvantages. • We define a generic architecture for TRS, identifying the main entities and processes involved. • We define a generic security framework for TRS. We identify the main security assets and propose a complete taxonomy of attacks for TRS. • We propose and validate a methodology to analyze, design, secure and deploy TRS in real-life environments. Additionally we identify the principal kind of applications we can implement with TRS and how TRS can provide a specific functionality. • We develop a software component to validate and optimize the behavior of a TRS in order to achieve a specific functionality or performance. In addition to the contributions made directly to the field of the TRS, we have made original contributions to different areas of knowledge thanks to the application of the analysis, design and security methodologies previously presented: • Detection of thermal anomalies in Data Centers. Thanks to the application of the TRS analysis and design methodologies, we successfully implemented a thermal anomaly detection system based on a TRS.We compare the detection performance of Self-Organized- Maps and Growing Neural Gas algorithms. We show how SOM provides better results for Computer Room Air Conditioning anomaly detection, yielding detection rates of 100%, in training data with malfunctioning sensors. We also show that GNG yields better detection and isolation rates for workload anomaly detection, reducing the false positive rate when compared to SOM. • Improving the performance of a harvesting system based on swarm computing and social odometry. Through the implementation of a TRS, we achieved to improve the ability of coordinating a distributed network of autonomous robots. The main contribution lies in the analysis and validation of the incremental improvements that can be achieved with proper use information that exist in the system and that are relevant for the TRS, and the implementation of the appropriated trust algorithms based on such information. • Improving Wireless Mesh Networks security against attacks against the integrity, confidentiality or availability of data and communications supported by these networks. Thanks to the implementation of a TRS we improved the detection time rate against these kind of attacks and we limited their potential impact over the system. • We improved the security of Wireless Sensor Networks against advanced attacks, such as insider attacks, unknown attacks, etc. Thanks to the TRS analysis and design methodologies previously described, we implemented countermeasures against such attacks in a complex environment. In our experiments we have demonstrated that our system is capable of detecting and confining various attacks that affect the core network protocols. We have also demonstrated that our approach is capable of rapid attack detection. Also, it has been proven that the inclusion of the proposed detection mechanisms significantly increases the effort the attacker has to introduce in order to compromise the network. Finally we can conclude that, to all intents and purposes, this thesis offers a useful and applicable knowledge in real-life environments that allows us to maximize the performance of any system based on a TRS. Thus, we deal with the main deficiency of this discipline: the lack of a common and complete base of knowledge and the lack of a methodology for the development of TRS that allow us to analyze, design, secure and deploy TRS in a systematic way.

Aplicação de técnicas inteligentes com análise no domínio do tempo para reconhecimento de defeitos em motores de indução trifásicos

Os motores de indução trifásicos são os principais elementos de conversão de energia elétrica em mecânica motriz aplicados em vários setores produtivos. Identificar um defeito no motor em operação pode fornecer, antes que ele falhe, maior segurança no processo de tomada de decisão sobre a manutenção da máquina, redução de custos e aumento de disponibilidade. Nesta tese são apresentas inicialmente uma revisão bibliográfica e a metodologia geral para a reprodução dos defeitos nos motores e a aplicação da técnica de discretização dos sinais de correntes e tensões no domínio do tempo. É também desenvolvido um estudo comparativo entre métodos de classificação de padrões para a identificação de defeitos nestas máquinas, tais como: Naive Bayes, k-Nearest Neighbor, Support Vector Machine (Sequential Minimal Optimization), Rede Neural Artificial (Perceptron Multicamadas), Repeated Incremental Pruning to Produce Error Reduction e C4.5 Decision Tree. Também aplicou-se o conceito de Sistemas Multiagentes (SMA) para suportar a utilização de múltiplos métodos concorrentes de forma distribuída para reconhecimento de padrões de defeitos em rolamentos defeituosos, quebras nas barras da gaiola de esquilo do rotor e curto-circuito entre as bobinas do enrolamento do estator de motores de indução trifásicos. Complementarmente, algumas estratégias para a definição da severidade dos defeitos supracitados em motores foram exploradas, fazendo inclusive uma averiguação da influência do desequilíbrio de tensão na alimentação da máquina para a determinação destas anomalias. Os dados experimentais foram adquiridos por meio de uma bancada experimental em laboratório com motores de potência de 1 e 2 cv acionados diretamente na rede elétrica, operando em várias condições de desequilíbrio das tensões e variações da carga mecânica aplicada ao eixo do motor.

Aplicações híbridas entre sistemas multiagentes e técnicas de inteligência artificial para redes inteligentes de distribuição de energia elétrica

Os smart grids representam a nova geração dos sistemas elétricos de potência, combinando avanços em computação, sistemas de comunicação, processos distribuídos e inteligência artificial para prover novas funcionalidades quanto ao acompanhamento em tempo real da demanda e do consumo de energia elétrica, gerenciamento em larga escala de geradores distribuídos, entre outras, a partir de um sistema de controle distribuído sobre a rede elétrica. Esta estrutura modifica profundamente a maneira como se realiza o planejamento e a operação de sistemas elétricos nos dias de hoje, em especial os de distribuição, e há interessantes possibilidades de pesquisa e desenvolvimento possibilitada pela busca da implementação destas funcionalidades. Com esse cenário em vista, o presente trabalho utiliza uma abordagem baseada no uso de sistemas multiagentes para simular esse tipo de sistema de distribuição de energia elétrica, considerando opções de controle distintas. A utilização da tecnologia de sistemas multiagentes para a simulação é baseada na conceituação de smart grids como um sistema distribuído, algo também realizado nesse trabalho. Para validar a proposta, foram simuladas três funcionalidades esperadas dessas redes elétricas: classificação de cargas não-lineares; gerenciamento de perfil de tensão; e reconfiguração topológica com a finalidade de reduzir as perdas elétricas. Todas as modelagens e desenvolvimentos destes estudos estão aqui relatados. Por fim, o trabalho se propõe a identificar os sistemas multiagentes como uma tecnologia a ser empregada tanto para a pesquisa, quanto para implementação dessas redes elétricas.

Arquitetura computacional para redes orgânicas e heterogêneas: plano de controle do sistema operacional swarm.

Computational Swarms (enxames computacionais), consistindo da integração de sensores e atuadores inteligentes no nosso mundo conectado, possibilitam uma extensão da info-esfera no mundo físico. Nós chamamos esta info-esfera extendida, cíber-física, de Swarm. Este trabalho propõe uma visão de Swarm onde dispositivos computacionais cooperam dinâmica e oportunisticamente, gerando redes orgânicas e heterogêneas. A tese apresenta uma arquitetura computacional do Plano de Controle do Sistema Operacional do Swarm, que é uma camada de software distribuída embarcada em todos os dispositivos que fazem parte do Swarm, responsável por gerenciar recursos, definindo atores, como descrever e utilizar serviços e recursos (como divulgá-los e descobrí-los, como realizar transações, adaptações de conteúdos e cooperação multiagentes). O projeto da arquitetura foi iniciado com uma revisão da caracterização do conceito de Swarm, revisitando a definição de termos e estabelecendo uma terminologia para ser utilizada. Requisitos e desafios foram identificados e uma visão operacional foi proposta. Esta visão operacional foi exercitada com casos de uso e os elementos arquiteturais foram extraídos dela e organizados em uma arquitetura. A arquitetura foi testada com os casos de uso, gerando revisões do sistema. Cada um dos elementos arquiteturais requereram revisões do estado da arte. Uma prova de conceito do Plano de Controle foi implementada e uma demonstração foi proposta e implementada. A demonstração selecionada foi o Smart Jukebox, que exercita os aspectos distribuídos e a dinamicidade do sistema proposto. Este trabalho apresenta a visão do Swarm computacional e apresenta uma plataforma aplicável na prática. A evolução desta arquitetura pode ser a base de uma rede global, heterogênea e orgânica de redes de dispositivos computacionais alavancando a integração de sistemas cíber-físicos na núvem permitindo a cooperação de sistemas escaláveis e flexíveis, interoperando para alcançar objetivos comuns.

A formal framework for modelling complex network management systems

Society today is completely dependent on computer networks, the Internet and distributed systems, which place at our disposal the necessary services to perform our daily tasks. Subconsciously, we rely increasingly on network management systems. These systems allow us to, in general, maintain, manage, configure, scale, adapt, modify, edit, protect, and enhance the main distributed systems. Their role is secondary and is unknown and transparent to the users. They provide the necessary support to maintain the distributed systems whose services we use every day. If we do not consider network management systems during the development stage of distributed systems, then there could be serious consequences or even total failures in the development of the distributed system. It is necessary, therefore, to consider the management of the systems within the design of the distributed systems and to systematise their design to minimise the impact of network management in distributed systems projects. In this paper, we present a framework that allows the design of network management systems systematically. To accomplish this goal, formal modelling tools are used for modelling different views sequentially proposed of the same problem. These views cover all the aspects that are involved in the system; based on process definitions for identifying responsible and defining the involved agents to propose the deployment in a distributed architecture that is both feasible and appropriate.

Modeling and exploiting behavior patterns in dynamic environments

This paper presents a new approach to improving the effectiveness of autonomous systems that deal with dynamic environments. The basis of the approach is to find repeating patterns of behavior in the dynamic elements of the system, and then to use predictions of the repeating elements to better plan goal directed behavior. It is a layered approach involving classifying, modeling, predicting and exploiting. Classifying involves using observations to place the moving elements into previously defined classes. Modeling involves recording features of the behavior on a coarse grained grid. Exploitation is achieved by integrating predictions from the model into the behavior selection module to improve the utility of the robot's actions. This is in contrast to typical approaches that use the model to select between different strategies or plays. Three methods of adaptation to the dynamic features of the environment are explored. The effectiveness of each method is determined using statistical tests over a number of repeated experiments. The work is presented in the context of predicting opponent behavior in the highly dynamic and multi-agent robot soccer domain (RoboCup)

The effect of social influence and curfews on civil violence

We investigate the policies of (1) restricting social influence and (2) imposing curfews upon interacting citizens in a community. We compare and contrast their effects on the social order and the emerging levels of civil violence. Influence models have been used in the past in the context of decision making in a variety of application domains. The policy of curfews has been utilised with the aim of curbing social violence but little research has been done on its effectiveness. We develop a multi-agent-based model that is used to simulate a community of citizens and the police force that guards it. We find that restricting social influence does indeed pacify rebellious societies, but has the opposite effect on peaceful ones. On the other hand, our simple model indicates that restricting mobility through curfews has a pacifying effect across all types of society.

Planning and decision-making for energy recovery from waste at the sub-regional level

This research project has developed a novel decision support system using Geographical Information Systems and Multi Criteria Decision Analysis and used it to develop and evaluate energy-from-waste policy options. The system was validated by applying it to the UK administrative areas of Cornwall and Warwickshire. Different strategies have been defined by the size and number of the facilities, as well as the technology chosen. Using sensitivity on the results from the decision support system, it was found that key decision criteria included those affected by cost, energy efficiency, transport impacts and air/dioxin emissions. The conclusions of this work are that distributed small-scale energy-from-waste facilities score most highly overall and that scale is more important than technology design in determining overall policy impact. This project makes its primary contribution to energy-from-waste planning by its development of a Decision Support System that can be used to assist waste disposal authorities to identify preferred energy-from-waste options that have been tailored specifically to the socio-geographic characteristics of their jurisdictional areas. The project also highlights the potential of energy-from-waste policies that are seldom given enough attention to in the UK, namely those of a smaller-scale and distributed nature that often have technology designed specifically to cater for this market.

Evolution of competing strategies in a threshold model for task allocation

A nature inspired decentralised multi-agent algorithm is proposed to solve a problem of distributed task allocation in which cities produce and store batches of different mail types. Agents must collect and process the mail batches, without global knowledge of their environment or communication between agents. The problem is constrained so that agents are penalised for switching mail types. When an agent process a mail batch of different type to the previous one, it must undergo a change-over, with repeated change-overs rendering the agent inactive. The efficiency (average amount of mail retrieved), and the flexibility (ability of the agents to react to changes in the environment) are investigated both in static and dynamic environments and with respect to sudden changes. New rules for mail selection and specialisation are proposed and are shown to exhibit improved efficiency and flexibility compared to existing ones. We employ a evolutionary algorithm which allows the various rules to evolve and compete. Apart from obtaining optimised parameters for the various rules for any environment, we also observe extinction and speciation.

Random distributed feedback fibre lasers

The concept of random lasers exploiting multiple scattering of photons in an amplifying disordered medium in order to generate coherent light without a traditional laser resonator has attracted a great deal of attention in recent years. This research area lies at the interface of the fundamental theory of disordered systems and laser science. The idea was originally proposed in the context of astrophysics in the 1960s by V.S. Letokhov, who studied scattering with "negative absorption" of the interstellar molecular clouds. Research on random lasers has since developed into a mature experimental and theoretical field. A simple design of such lasers would be promising for potential applications. However, in traditional random lasers the properties of the output radiation are typically characterized by complex features in the spatial, spectral and time domains, making them less attractive than standard laser systems in terms of practical applications. Recently, an interesting and novel type of one-dimensional random laser that operates in a conventional telecommunication fibre without any pre-designed resonator mirrors-random distributed feedback fibre laser-was demonstrated. The positive feedback required for laser generation in random fibre lasers is provided by the Rayleigh scattering from the inhomogeneities of the refractive index that are naturally present in silica glass. In the proposed laser concept, the randomly backscattered light is amplified through the Raman effect, providing distributed gain over distances up to 100km. Although an effective reflection due to the Rayleigh scattering is extremely small (~0.1%), the lasing threshold may be exceeded when a sufficiently large distributed Raman gain is provided. Such a random distributed feedback fibre laser has a number of interesting and attractive features. The fibre waveguide geometry provides transverse confinement, and effectively one-dimensional random distributed feedback leads to the generation of a stationary near-Gaussian beam with a narrow spectrum. A random distributed feedback fibre laser has efficiency and performance that are comparable to and even exceed those of similar conventional fibre lasers. The key features of the generated radiation of random distributed feedback fibre lasers include: a stationary narrow-band continuous modeless spectrum that is free of mode competition, nonlinear power broadening, and an output beam with a Gaussian profile in the fundamental transverse mode (generated both in single mode and multi-mode fibres).This review presents the current status of research in the field of random fibre lasers and shows their potential and perspectives. We start with an introductory overview of conventional distributed feedback lasers and traditional random lasers to set the stage for discussion of random fibre lasers. We then present a theoretical analysis and experimental studies of various random fibre laser configurations, including widely tunable, multi-wavelength, narrow-band generation, and random fibre lasers operating in different spectral bands in the 1-1.6μm range. Then we discuss existing and future applications of random fibre lasers, including telecommunication and distributed long reach sensor systems. A theoretical description of random lasers is very challenging and is strongly linked with the theory of disordered systems and kinetic theory. We outline two key models governing the generation of random fibre lasers: the average power balance model and the nonlinear Schrödinger equation based model. Recently invented random distributed feedback fibre lasers represent a new and exciting field of research that brings together such diverse areas of science as laser physics, the theory of disordered systems, fibre optics and nonlinear science. Stable random generation in optical fibre opens up new possibilities for research on wave transport and localization in disordered media. We hope that this review will provide background information for research in various fields and will stimulate cross-disciplinary collaborations on random fibre lasers. © 2014 Elsevier B.V.

Optimising market share and profit margin:SMDP-based tariff pricing under the smart grid paradigm

Smart grid technologies have given rise to a liberalised and decentralised electricity market, enabling energy providers and retailers to have a better understanding of the demand side and its response to pricing signals. This paper puts forward a reinforcement-learning-powered tool aiding an electricity retailer to define the tariff prices it offers, in a bid to optimise its retail strategy. In a competitive market, an energy retailer aims to simultaneously increase the number of contracted customers and its profit margin. We have abstracted the problem of deciding on a tariff price as faced by a retailer, as a semi-Markov decision problem (SMDP). A hierarchical reinforcement learning approach, MaxQ value function decomposition, is applied to solve the SMDP through interactions with the market. To evaluate our trading strategy, we developed a retailer agent (termed AstonTAC) that uses the proposed SMDP framework to act in an open multi-agent simulation environment, the Power Trading Agent Competition (Power TAC). An evaluation and analysis of the 2013 Power TAC finals show that AstonTAC successfully selects sell prices that attract as many customers as necessary to maximise the profit margin. Moreover, during the competition, AstonTAC was the only retailer agent performing well across all retail market settings.

A New Class of Processes for Formalizing and Generalizing Individual-Based Models: The Semi-Semi-Markov Processes

2000 Mathematics Subject Classification: 60K15, 60K20, 60G20,60J75, 60J80, 60J85, 60-08, 90B15.