Basal cell carcinoma is a form of a non-melanoma skin tumour, that commonly forms over the sun exposed regions of the head and neck. Investigation of the rate of occurrence at different sites on face and neck shows considerable variation from site to site. The inner canthus has a disproportionate number when compared to more exposed sites. The eye brow ridge, cheek bone and nose limit the field of view of the inner canthus, thus it is expected to receive less radiation than other more exposed regions. To explain the disproportionate rate, it is hypothesised that a portion of radiation incident onto the eye is reflected to the inner canthus. The aim of this thesis is to investigate the contribution that the radiation reflected off the surface of the eye makes to the overall dose on the inner canthus. The inter reflections between the eye and inner canthus were studied through the use of the ray tracing program Zemax. Zemax was used to trace rays in a non sequential mode incident onto a model eye and periorbital region. To obtain the models of the eye and periorbital region, both magnetic resonance imaging and a casting process was investigated, with the later being superior for our uses. With the model obtained, it was used in a series of three dimensional ray tracing programs. On a macroscopic scale there is a small increase in the irradiance on the inner canthus (2 % over a 1 cm2 area). Peaks of high irradiance (19 % increase in irradiance above direct irradiance) were discovered over the surface when the detector was divided into 200 mm elements. It was concluded that these increases above the direct irradiance in these small regions, increases the possibility of the occurrence of a Basal cell carcinoma. Individual facial geometry, will greatly effect the location and size of these peaks and as a result an experimental method to measure the dose distribution across the inner canthus was proposed. Initially it was planned to use polysulphone film to measure the erythemal dose on the inner canthus. Results from the modelling indicated that any measurements made had to be at a high spatial resolution. Polysulphone film was found to be inadequate for this, due to its large uncertainties. An alternative method was investigated so that a population study could be performed in future studies using visible radiation and high dynamic range images gave a simple and effective clinical assessment tool. The high dynamic range images showed hot spots in the irradiance across the inner canthus agreeing with the model. The small spots of high relative irradiance may not be the only reason for the increased rate in this region. Greater skin sensitivity and absence of sun screen use at this site are other possibilities. It is believed however that the irradiance distribution across the inner canthus on a microscopic scale goes a long way to increasing the risk for certain people.

This review has examined the spatial and temporal neural activation of speech comprehension. Six theories on speech comprehension were selected and reviewed. The most fundamental structures for speech comprehension are the superior temporal gyrus, the fusiform gyrus, the temporal pole, the temporoparietal junction, and the inferior frontal gyrus. Considering temporal aspects of processes, the N400 ERP effect indicates semantic violations, and the P600 indicates re-evaluation of a word due to ambiguity or syntax error. The dual-route processing model provides the most accurate account of neural correlates and streams of activation necessary for speech comprehension, while also being compatible with both the reviewed studies and the reviewed theories. The integrated theory of language production and comprehension provides a contemporary theory of speech production and comprehension with roots in computational neuroscience, which in conjunction with the dual-route processing model could drive the fields of language and neuroscience even further forward.

Cette thèse d'ingénierie des connaissances a pour objet la redocumentation de l'activité médiée informatiquement à base de fichiers traces. Au cours d’un tel processus, un utilisateur interprète et réécrit les traces dans un nouveau document, qui correspond à un genre particulier et décrit l'activité dans toute sa richesse ; il a notamment la possibilité de reformuler le contenu des traces, de l'enrichir avec d'autres informations (justification ou informations sur le contexte), de supprimer des parties, etc. Il s’agit ainsi de renforcer le statut documentaire des traces selon les trois dimensions de la lisibilité, de l’intelligibilité et de l’échange. Nous proposons une approche générique semi-automatique qui combine le contenu de traces de l’activité et les connaissances des utilisateurs dans des formes narratives, facilement échangeables, décrivant l’activité de façon personnalisée pour les utilisateurs. Celle-ci utilise des modèles formels pour les traces et les documents produits, ainsi que les principes de la théorie de la structure rhétorique (RST). Elle soutient la redocumentation à travers deux phases : une phase automatique pour générer un document initial fragmenté à partir de la trace d'activité comme une première description pour celle-ci et une phase interactive pour personnaliser cette description selon les besoins et les choix d'un utilisateur afin de produire un document. Nous proposons également la spécification de cette approche pour le cas particulier de la redocumentation par le texte et nous présentons ActRedoc, un outil auteur que nous avons développé pour implémenter cette approche spécifique en utilisant différentes technologies du Web Sémantique. Les résultats de notre travail ont été testés dans un environnement réel qui est celui de la redocumentation des traces de navigation dans un site d’archives de journaux du XIXème développé par la bibliothèque municipale de Lyon. Nous présentons ainsi les résultats d'une première évaluation de l'approche de redocumentation en texte et de l'outil ActRedoc auprès d'un groupe d'utilisateurs / This thesis is situated within the framework of knowledge engineering; it is particularly interested in the redocumentation of the computermediated activity from traces files. We suppose that during such a process, a user interprets and rewrites traces in a new document that corresponds to a particular genre and describes the traced activity in détails ; all this via the possibility to rewrite traces content, to enrich this content with other information (justifications or information on the context), to delete parts from it, etc. Therefore, we strengthen the documentary status of traces according to the three dimensions of legibility, comprehensibility and exchange. For implementing such a process, we propose a generic semi-automatic approach which combines the content of activity traces with users knowledge in narrative forms, easily exchangeable, to describe the traced activity in a personalized way for users. This Approach uses formal models for traces and documents (produced), as well as principles of the rhetorical structure theory (RST). It supports the redocumentation process through two phases: an automatic phase to generate an initial fragmented document from traces as a first description of the activity and an interactive phase to personalize this description according to the user needs and choices so that to produce an easily exchangeable document. We also propose a specification of this approach for the particular case of redocumentation by text and we present ActRedoc, an authoring tool that we developed by using different technologies of the Semantic Web in order to implement this specific approach. The results of our work were tested in a real environment which concerns redocumenting the activity of browsing a site of archives of newspapers of the XIXth century developed by the municipal library of Lyon. Thus, we present a first evaluation of the approach of redocumentation by text and of the authoring tool ActRedoc by a group of users