Esta investigación busca encajar la obra de Román Gil en la tradición del libro de artista. En una época en la que éste se encuentra en su momento de madurez, él mismo ha posibilitado el que tanto una imagen de lugar a su verbalización como el que un texto escrito devenga en una imagen visual. Ahora es también el momento de incorporar a esta incipiente disciplina parte de la genealogía que ha deparado la configuración del libro como lugar de todos los elementos temporales que constituyen la lengua hablada. Así visto, el libro se manifiesta como una forma simbólica autónoma y bien distinta que el texto al que da sustento. Es en este punto donde hemos querido comprobar si en la obra de Gil se da ese valor genuino del arte capaz de hacernos aprehender los significados de las imágenes a través de la forma. El factor de regulación al que queremos aludir y que goza de los valores gráficos, escultóricos y pictóricos de todo el itinerario que dio lugar a la configuración de los sistemas escritos, es el que sospechamos que implícitamente subyace, más que en la escritura, en la configuración de ésta. Paradójicamente, es en la obra de un autor que hace libros de artista con una expresa renuncia a la palabra escrita, donde buscamos la hermenéutica de los símbolos que han conseguido objetivar la realidad con la mayor precisión habida a nuestro alcance. Pero no es para comprenderla bajo este punto de vista, sino para averiguar si las imágenes plásticas son capaces de ser elocuentes a través de un objeto -el libro- que ha sido concebido como sede espacial de la contingencia temporal que es el habla. De la mano de la obra de Gil conocida como las Carpetas de Román, nos hemos adentrado en la prehistoria de la escritura para tratar de entender cómo se acaba encriptando en ella todo el ritual de la experiencia plástica y cómo en este proceso surge un objeto tan específico como pueda ser el libro que, para nosotros está al mismo nivel conceptual que el cuadro. Más tarde, hemos observado el proceso de creación de este artista para averiguar de qué manera consigue transferir las particularidades del género pictórico del autorretrato a la disciplina del libro de artista, sabiendo que este autor renuncia a la representación formal de los datos plásticos emparentados con el parecido anatómico y psicológico. Hemos investigado también su capacidad para referirse a la realidad a través de los órganos sensoriales regentes de parcelas sensitivas de las que no dan debida cuenta y hemos tratado de averiguar si las imágenes que detrae de estas experiencias son vividas o dependen de un carácter metafórico de carácter cultural. Posteriormente, indagamos en el tipo de formación académica y las afinidades electivas que hayan podido influir en su talante y en su voluntad de hacer un arte a través de un medio que parece haber surgido para reflejar la materialización perceptual de la realidad a través de la palabra. A partir de este punto de vista, hemos podido comprobar cómo Gil actualiza el modelo canónico que dio lugar a las primeras bibliotecas a través de las series artísticas que emprende y en base a los ordenamientos que establece con su propia producción artística. Por último, hemos analizado su quehacer artístico, así como los materiales empleados en su obra con la intención de aprehender su manera de unir géneros pictóricos con materiales experimentales y procedimientos alternativos, más allá de sus usos y de sus funciones. / This research is an attempt to place the work of Román Gil within the artist book tradition. In his maturity stage, he himself has made it possible for an image to become its verbal representation, as well as enable a written text to become a visual image. Now it is also the time to incorporate to this incipient discipline, part of the genealogy which has turned the structure of a book into the meeting place of all the temporal elements constituting the spoken languages. It is at this point that we have sought to prove if, in Gil's work, there can be found that genuine artistic value which can allow us to grasp the meanings of the imagery through its form. The control factor to which we want to refer, and which contains the graphic, sculptural and pictorial values of the whole itinerary originating the structure of the written systems, is which we tend to believe that implicitly underlies, rather than writing, its layout. Paradoxically, it is in the work of an author creating artist books and explicitly renouncing the written word, where we look for the hermeneutics of those symbols which have attained to make reality tangible, with the greatest possible precision available to us. But it is not our aim to understand it from his point of view but to enquire if the plastic images can express themselves through an object, -the book-, which has been conceived as the place for the temporal contingency given in speech. Hand in hand with Gil's work, known as the Román Files, we have ventured into the prehistory of writing, in an attempt to understand how it is possible to arrive at encrypting in it all the ritual of the plastic arts experience, and how, in the process, such a specific object as a book, which to us has the same conceptual level as a picture, takes on life. Furthermore, we have observed this artist's creative process to discover how he manages to transfer the specifics of the self-portrait pictorial genre to the artist book discipline, considering that this author renounces the formal representation of the plastic arts data related to the anatomic and psychological resemblance. We have also examined his capacity to refer to reality through the sensory organs governing unaccounted for sensitive areas, aiming to discover if the images he draws from these experiences come from life or derive from a metaphorical cultural character. Later still, we enquired into his academic background and his selective affinities which could have influenced his character, and his decision to make an art out of a medium which seems to have been created to reflect the perceptual materialization of reality through the word. From this standpoint, we have been able to understand how Gil brings up to date the canonical model which originated the first libraries to appear, through the artistic series he produces and on the basis of the order he sets on his own artwork. Lastly, we have analyzed his production as well as the materials he uses, in an attempt to grasp his way of combining pictorial genres by means of experimental materials and alternative procedures, which go beyond their customary uses and functions.

This thesis investigates the distinct contributions of artists who engage the public with climate data, exploring the role of the artists and the value of their contributions by focusing on two studies of artist-led projects. The first investigates how the author’s own artist led collective, Active Ingredient, engaged the public with climate data through a touring interactive artwork. The second study explores the design of an online platform for capturing, authoring and ‘performing’ climate data. This platform was developed and tested by Active Ingredient in collaboration with researchers from the University of Nottingham, and then used by other artists to engage with climate data. The studies reveal how the artistic projects were designed and experienced, through a mixed methods approach requiring the author to shift perspectives in order to investigate her own arts practice alongside the work of other artists in this field. The findings from these studies suggest that the artists adopt a distinctive voice that fosters an emotional engagement with climate data, rather than an informative or persuasive one, that goes beyond ‘environmental knowing’ towards human-scale, embodied, localized and personalized sense making. This research reveals how the artists use the key strategies of performing data, sensory experience and multiple interpretations to provoke these emotional responses. Highlighting the challenges and opportunities of engaging temporal structures and narratives to represent climate data; treating the data as a new material that is embedded into the artworks and embodied in various sensory forms; abstracting and juxtaposing multiple, contrasting and yet related datasets so as to invite comparisons, while opening up spaces between them for interpretation and dialogue. This results in a discussion of the role of technology within the artistic process, how the artists walk a line between authenticity and emotional engagement in their interpretations of climate data and the importance of an ongoing dialogic collaboration between the artists, researchers and climate scientist that support authentic and meaningful engagements with climate data. The research presents rich descriptions of the artists’ strategies for engaging the public with climate data and revealing that artists have a distinctive and powerful role to play in relation to climate change and sustainability; one that Human Computer Interaction (HCI) and Climate Science need to understand as they continue to move into this territory, and where HCI in particular might ultimately learn about how to bring an emotional treatment to many other forms of data.

The novel contribution of this research is the development of a supervised algorithm that extracts relevant attributes from The Health Improvement Network database to detect prescription side effects. Prescription drug side effects are a common cause of morbidity throughout the world. Methods that aim to detect side effects have historically been limited due to the data available, but some of these limitations may be overcome by incorporating longitudinal observational databases into pharmacovigilance. Existing side effect detecting methods using longitudinal observational databases have shown promise at becoming a fundamental component of post marketing surveillance but unfortunately have high false positive rates. An extra step is required to further analyse and filter the potential side effects detected by existing methods due to their high false positive rates, and this reduces their efficiency. In this thesis a novel methodology, the supervised adverse drug reaction predictor (SAP) framework, is presented that learns from known side effects, and identifies patterns that can be utilised to detect unknown side effects. The Bradford-Hill causality considerations are used to derive suitable attributes as inputs into a learning algorithm. Both supervised and semi-supervised techniques are investigated due to the limited number of definitively known side effects. The results showed that the SAP framework implementing a random forest classifier outperformed the existing methods on The Health Improvement Network longitudinal observational database, with AUCs ranging between 0.812-0.937, an overall MAP of 0.667, precision values between 0.733-1 and a false positive rate ≤ 0.013. When applied to the standard reference the SAP framework implementing a support vector machine obtained a MAP score of 0.490, an average AUC of 0.703 and a false positive rate of 0.16. The false positive rate is lower than that obtained by existing methods on the standard reference.

Examining patient records is a useful way to identify common conditions and treatment outcomes in veterinary practice and data gathered can be fed back to the profession to assist with clinical decision making. This research aimed to develop a method to extract clinical data from veterinary electronic patient records (EPRs) and to assess the value of the data extracted for use in practice-based research. The transfer of new research from continuing professional development (CPD) into practice was also considered. An extensible mark-up language (XML) schema was designed to extract information from a veterinary EPR. The analysis of free text was performed using a content analysis program and a clinical terms dictionary was created to mine the extracted data. Data collected by direct observation was compared to the extracted data. A review of research published in the proceedings of a popular veterinary CPD event, British Small Animal Veterinary Association (BSAVA) Congress, was appraised for evidence quality. All animal records were extracted and validation confirmed 100% accuracy. The content analysis produced results with a high specificity (100%) and the mined data analysis was successful in assessing the prevalence of a specific disease. On comparison, the data extracted from the EPR contained only 65% of all data recorded by direct observation. The review of BSAVA Congress abstracts found the majority of the clinical research abstracts (CRAs) presented to be case reports and case series, with differences in focus between CRAs and veterinary lecture stream abstracts. This study has demonstrated that data extraction using an XML schema is a viable method for the capture of patient data from veterinary EPRs. The next step will be to understand the differences found between data collected by observation and extraction, and to investigate how research presented as CPD is received, appraised and applied by the veterinary profession.

Stroke is a major cause of physical disability. Recovery is possible, and can continue indefinitely. As such, much of it will take place at home, often with minimal support from professional therapists. As computing becomes more pervasive and familiar, opportunities exist to design technology to support rehabilitation in the home environment. However, given the varied nature of disabilities caused by stroke, there is a need for a greater understanding of how to design technology that is sufficiently tailored to the needs of individuals and which is appropriate for usage in their homes. This thesis offers an exploration of these issues, through a series of research activities constructed around the direct participation of stroke survivors and their families. The core of this thesis begins with a consideration of a focus group which was attended by survivors of stroke and their partners. Recorded discussions provide a rich insight into their collective experience of living with stroke, and the implications of these findings for the design of effective rehabilitation technologies are considered. The design of bespoke technologies which were directly tailored to the rehabilitational needs and personal motivations of four stroke survivors is then described. Prototypes of these technologies were deployed for periods ranging from one to seven months. Data recorded throughout this entire process provides a detailed understanding of the factors that have influenced their design, use and impact. Through an analysis of material collected during all of these engagements, this thesis presents a set of contributions which can support the design of better home-based rehabilitation technologies in the future. These contributions support a more general understanding of the interactional needs of individuals who have experienced a dramatic and potentially traumatic change in their life, and of mechanisms for tailoring persuasive computing technologies to the specific motivations of those who use them.

Vehicle Satellite Navigation Systems are commonplace but from a safety point of view, can be problematic. They are situated on small screens and often placed at an angle which is not in the driver’s field of view. This promotes glances away from the road and reduces situational awareness for events which occur. Audio commands and visual maps which are used to instruct drivers typically require translation to situationally meaningful actions before execution in the real world. This increases the driver’s workload and risk of distraction. A virtual car head-up display concept which is novel to this thesis is introduced. The design was motivated by issues in the literature regarding workload and risk of distraction and was shaped using a field study. Also, as head-up displays are becoming common in new vehicles, the benefits they offer have been explored in the design of the virtual car head-up display. Navigation instructions would be embedded in a familiar object to the driver; a car image, to support driving practices (e.g. indicating, lane positioning and turning) which are absent in the abstract commands and visual maps employed by vehicle satellite navigation systems. The navigation instructions used by the virtual car head-up display are easy to understand and can reduce the processing times for the instructions. For example, rather than translate audio commands e.g. “after 200 yards turn left”, the driver sees the virtual car indicate left 200 yards from the turn and sees it turn left on arrival at the turn. Also, rather than translate complex visual maps, the driver replicates the actions of the virtual car. An initial prototype for the virtual car head-up display was designed after which usability evaluation was carried out in a driving simulator to refine the concept. The first two studies were part of the design process and involved assessing the feasibility and conformity of the virtual car head-up display. It was found that the virtual car head-up display was an intelligible way to present the navigation instructions to drivers and that it was better to conform the virtual car to the external environment. The third study compared the prototype of the virtual car head-up display with the prototype of an arrow head-up display and vehicle satellite navigation system. It was found that the virtual car head-up display had the least workload and risk of distraction and was the easiest to use. A synthesis of the research work is provided which outlines the key contributions to research.

Κυριότερος σκοπός είναι η αξιολόγηση των συστημάτων Υπολογιστικής Τομογραφίας και Μαγνητικού Συντονισμού κατά την εφαρμογή της τεχνικής Δυναμικής Αιμάτωσης (Perfusion). Επίσης η δημιουργία και εφαρμογή αναλυτικού και αποτελεσματικού πρωτοκόλλου στην οξεία ισχαιμία με σκοπό την άμεση βοήθεια του ισχαιμικού εγκεφάλου. / Thesis MSc in order to highlight the usefulness of CT Brain Perfusion in acute brain ischemia.

Δεδομένης και της χρήσης της τεχνικής SPECT, ο σκοπός της παρούσας διπλωματικής εργασίας είναι η διερεύνηση του κατά πόσο μπορούν, συνδυαστικά, οι 5 αυτές τεχνικές να δείξουν μία πιο σαφή εικόνα στη διαφορική διάγνωση όγκων του ΚΝΣ. Συνδυάζοντας τα δεδομένα που θα ληφθούν από τις μετρήσεις θα δημιουργηθεί μια μικρή βάση δεδομένων η οποία ως απώτερο στόχο θα έχει τη διεύρυνσή της στο μέλλον και συνεπώς την εξαγωγή ασφαλέστερων συμπερασμάτων. / Combination and evaluation of advanced MR techniques (MRS, DWI, DTI, PWI) and Scintigraphy in the differential diagnosis of Central Nervous System tumors.

「華航事件」發生於民國七十五年五月，是海峽兩岸對立近四十年首次正式接觸，雖 然事件的解決限定於民間、商業的性質，但不可否認的，此事件帶給中華民國不小的 衝擊，以「危機」視之並不為過。雖然結果順利解決，但實為兩岸間關係的突破種下 契機，是以「華航事件」的研究深具意義。 本論文擬採理論與實例並重的研究方法，以理論為架構，實例為印證。但是，理論部 份僅作概念性介紹，華航事件的研究才是本論文之主要旨趣。換言之，以「危機管理 」的三個時期（爆發、解決、善後）分層剖析華航事件，最後提出檢討與結論。 第一章為導論，闡述本論文之研究動機、架構、範圍及方法。第二章探討危機之界定 與理論之運用。第三章概論危機管理的理論。 第四章以後研究華航事件，第四章為背景介紹。第五章為危機爆發期，第六章為危機 解決期，第七章為善後處理期，第八章則對華航事件提出檢討，第九章為結論。希望 本論文的研究對危機管理理論與實務，以及兩岸關係有所啟示。

Despite over 30 years of research, face recognition is still one of the most difficult problems in the field of Computer Vision. The challenge comes from many factors affecting the performance of a face recognition system: noisy input, training data collection, speed-accuracy trade-off, variations in expression, illumination, pose, or ageing. Although relatively successful attempts have been made for special cases, such as frontal faces, no satisfactory methods exist that work under completely unconstrained conditions. This thesis proposes solutions to three important problems: lack of training data, speed-accuracy requirement, and unconstrained environments. The problem of lacking training data has been solved in the worst case: single sample per person. Whitened Principal Component Analysis is proposed as a simple but effective solution. Whitened PCA performs consistently well on multiple face datasets. Speed-accuracy trade-off problem is the second focus of this thesis. Two solutions are proposed to tackle this problem. The first solution is a new feature extraction method called Compact Binary Patterns which is about three times faster than Local Binary Patterns. The second solution is a multi-patch classifier which performs much better than a single classifier without compromising speed. Two metric learning methods are introduced to solve the problem of unconstrained face recognition. The first method called Indirect Neighourhood Component Analysis combines the best ideas from Neighourhood Component Analysis and One-shot learning. The second method, Cosine Similarity Metric Learning, uses Cosine Similarity instead of the more popular Euclidean distance to form the objective function in the learning process. This Cosine Similarity Metric Learning method produces the best result in the literature on the state-of-the-art face dataset: the Labelled Faces in the Wild dataset. Finally, a full face verification system based on our real experience taking part in ICPR 2010 Face Verification contest is described. Many practical points are discussed.