Ce travail de thèse concerne le traitement d’image fDOT (fDOT pour fluorescence diffuse optical tomography) suit vers deux axes. Le recalage d'images fDOT à l’aide de l’imagerie TEP (tomographie par émission de positons) et l’amélioration des reconstructions fDOT à l’aide de miroirs pour collecter des projections complémentaires. Il est présenté en deux parties : Dans la première partie, une méthode automatique pour recaler les images de fDOT avec les images de Tomographie par Emission de Positons (TEP) développée dans le but de corréler l’ensemble des informations issues de chaque modalité. Cette méthode de recalage est basée sur une détection automatique de marqueurs fiduciaires présents dans les deux modalités. La particularité de cette méthode est l’utilisation de l’image de surface obtenue en fDOT, qui sert à identifier la position en Z des marqueurs fiduciaires dans les images optiques. Nous avons testé cette méthode sur un modèle de souris porteuses de xénogreffes de tumeurs de cellules cancéreuses MEN2A qui imitent un carcinome thyroïdien médullaire humain, après une double injection de traceur radioactif : [18F]2-fluoro-2-Deoxy-D-glucose (FDG) pour l’imagerie TEP et un traceur optique d’infrarouge fluorescent, le Sentidye. Grâce à la précision de notre méthode, nous arrivons à démontrer que le signal Sentidye est présent à la fois dans la tumeur et les vaisseaux environnants [1]. La qualité des images fDOT est dégradée selon l’axe Z du fait d’un nombre limité de projections pour la reconstruction. Dans la deuxième partie, le travail s’est orienté vers une nouvelle méthode de reconstruction d’images fDOT à partir d’un nouveau système d’acquisition multi-angulaire avec deux miroirs placés de chaque côté de l’animal. Ce travail a été mené en collaboration avec le département CS d’University College London (UCL), partenaire du projet Européen FMT-XCT. Le logiciel TOAST développé par cette équipe a été utilisé comme source pour l’algorithme de reconstruction, et modifié pour s’adapter à notre problématique. Après plusieurs essais concernant l’ajustement des paramètres du programme, nous avons appliqué cette méthode sur un fantôme réaliste des tissus biologiques et chez la souris. Les résultats montrent une amélioration de l’image reconstruite d’un fantôme semi-cylindrique et de l’image de rein chez la souris, pour lesquelles la méthode des miroirs est supérieure à la méthode classique sans miroir. Malgré tout, nous avons observé que les résultats étaient très sensibles à certains paramètres, d’où une performance de reconstruction variable d’un cas à l’autre. Les perspectives futures concernent l’optimisation des paramètres afin de généraliser l’approche multi-angle. / This thesis concerns the image processing of fluorescence diffuse optical tomography (fDOT), following two axes: FDOT image co-registration with PET (positron emission tomography) image and improvement of fDOT image reconstructions using mirrors to collect additional projections. It is presented in two parts:In the first part, an automatic method to co-register the fDOT images with PET images has been developed to correlate all the information from each modality. This co-registration method is based on automatic detection of fiducial markers (FM) present in both modalities. The particularity of this method is the use of optical surface image obtained in fDOT imaging system, which serves to identify the Z position of FM in optical images. We tested this method on a model of mice bearing tumor xenografts of MEN2A cancer cells that mimic a human medullary thyroid carcinoma, after a double injection of radiotracer [18F] 2-fluoro-2-Deoxy-D-glucose ( FDG) for PET imaging and optical fluorescent infrared tracer Sentidye. With the accuracy of our method, we can demonstrate that the signal of Sentidye is present both in the tumor and surrounding vessels.The fDOT reconstruction image quality is degraded along the Z axis due to a limited number of projections for reconstruction. In the second part, the work is oriented towards a new method of fDOT image reconstruction with a new multi-angle data acquisition system in placing two mirrors on each side of the animal. This work was conducted in collaboration with the CS Department of University College London (UCL), a partner of the European project FMT-XCT. TOAST software developed by this team was used as source code for the reconstruction algorithm, and was modified to adapt to the concerned problem. After several tests on the adjustment of program parameters, we applied this method on a phantom that simulating the biological tissue and on mice. The results showed an improvement in the reconstructed image of a semi-cylindrical phantom and the image of mouse kidney, for which the reconstruction of the mirrors geometry is better than that of conventional geometry without mirror. Nevertheless, we observed that the results were very sensitive to certain parameters, where the performance of reconstruction varies from one case to another. Future prospectives concern the optimization of parameters in order to generalize the multi-angle approach.

Ce travail de thèse concerne le traitement d'image fDOT (fDOT pour fluorescence diffuse optical tomography) suit vers deux axes. Le recalage d'images fDOT à l'aide de l'imagerie TEP (tomographie par émission de positons) et l'amélioration des reconstructions fDOT à l'aide de miroirs pour collecter des projections complémentaires. Il est présenté en deux parties : Dans la première partie, une méthode automatique pour recaler les images de fDOT avec les images de Tomographie par Emission de Positons (TEP) développée dans le but de corréler l'ensemble des informations issues de chaque modalité. Cette méthode de recalage est basée sur une détection automatique de marqueurs fiduciaires présents dans les deux modalités. La particularité de cette méthode est l'utilisation de l'image de surface obtenue en fDOT, qui sert à identifier la position en Z des marqueurs fiduciaires dans les images optiques. Nous avons testé cette méthode sur un modèle de souris porteuses de xénogreffes de tumeurs de cellules cancéreuses MEN2A qui imitent un carcinome thyroïdien médullaire humain, après une double injection de traceur radioactif : [18F]2-fluoro-2-Deoxy-D-glucose (FDG) pour l'imagerie TEP et un traceur optique d'infrarouge fluorescent, le Sentidye. Grâce à la précision de notre méthode, nous arrivons à démontrer que le signal Sentidye est présent à la fois dans la tumeur et les vaisseaux environnants [1]. La qualité des images fDOT est dégradée selon l'axe Z du fait d'un nombre limité de projections pour la reconstruction. Dans la deuxième partie, le travail s'est orienté vers une nouvelle méthode de reconstruction d'images fDOT à partir d'un nouveau système d'acquisition multi-angulaire avec deux miroirs placés de chaque côté de l'animal. Ce travail a été mené en collaboration avec le département CS d'University College London (UCL), partenaire du projet Européen FMT-XCT. Le logiciel TOAST développé par cette équipe a été utilisé comme source pour l'algorithme de reconstruction, et modifié pour s'adapter à notre problématique. Après plusieurs essais concernant l'ajustement des paramètres du programme, nous avons appliqué cette méthode sur un fantôme réaliste des tissus biologiques et chez la souris. Les résultats montrent une amélioration de l'image reconstruite d'un fantôme semi-cylindrique et de l'image de rein chez la souris, pour lesquelles la méthode des miroirs est supérieure à la méthode classique sans miroir. Malgré tout, nous avons observé que les résultats étaient très sensibles à certains paramètres, d'où une performance de reconstruction variable d'un cas à l'autre. Les perspectives futures concernent l'optimisation des paramètres afin de généraliser l'approche multi-angle.

Citizens in the United States are fortunate to have an excellent system of roadways and the affluence with which to afford automobiles. The flexibility of travel on demand for most allows for a variety of lifestyles, assists with conducting business, and contributes to the feeling of freedom that most citizens enjoy. The current vehicle fleet, which is primarily powered by internal combustion engines burning fossil fuels, does however contribute to the deterioration of air quality. This effect is particularly significant in metropolitan areas. Motor vehicle exhausts contain several combustion bi-products that pose harmful effects to the environment and human health, in particular. The United States Environmental Protection Agency (EPA) and the Federal Highway Administration (FHWA) have selected carbon monoxide (CO) as the air pollutant on which it has based its guidelines for assessing potential air quality impacts from roadway construction (EPA 1992). The design of roadway networks must consider traffic flows, Level of Service (LOS), cost, and National Ambient Air Quality Standards (NAAQS) requirements. In light of the environmental standards it is necessary to model to estimate potential future near-road concentrations of CO. This modeling has two aspects, first determining the rate of pollutant emissions, and second determining how those pollutants disperse near the road. Obtaining a precise, realistic estimate of the near-road CO concentrations under a wide variety of weather and traffic patterns is a potentially huge undertaking. With budgetary constraints in mind, the development of a screening model is appropriate. CO Florida 2012 (COFL2012) is such a model that uses conservative assumptions to predict worst-case, near-road CO concentration. Projects that pass a COFL2012 model run do not require additional air quality modeling. Projects that fail a COFL2012 model run, however, may still be viable, but will require additional, detailed modeling and possibly project modifications. COFL2012 uses tables of emission factors (EFs) that were derived from numerous runs of the EPA's MOtor Vehicle Emission Simulator (MOVES2010a), which is indicated as the preferred model for near-road modeling of CO.(EPA 2009) COFL2012 then inputs the EFs, along with assumed link configurations, geographical assumptions, and user-inputted traffic information into input files that are run through CAL3QHC Version 2.0 (CAL3QHC2), the EPA's approved near-road dispersion model (EPA 1995). COFL2012 is a brand new Florida CO screening model, written from scratch. This author has written the computer code for COFL2012 in Visual Basic, using Microsoft Visual Studios 2010. Visual Studios utilizes the .net Framework 4. COFL2012 is easy to learn, quick to operate, and has been written to allow for future updates simply and easily, whenever the EPA releases updates to the databases that feed MOVES2010a.

The goal of this thesis is to evaluate travelers' experience with Highway Advisory Radio (HAR) and Citizens' Band Radio Advisory System (CBRAS) technologies on both Florida Interstate Highway system (FIH) and the Florida Turnpike Enterprise (FTE) toll roads. To achieve this goal, two different survey tools were used. The first tool is a random digit dialing phone survey known as CATI (Computer-Assisted Telephone Interviewing). The second tool is a field survey that intercepts travelers at the Florida Turnpike Enterprise (FTE) service plazas and the Florida Interstate Highway (FIH) rest areas. HAR and CBRAS are traditional components of the Advanced Traveler Information Systems (ATIS). This thesis pays special attention to the effectiveness of HAR and CBRAS in improving travelers' experience. Feedback to analyze these two technologies was collected via a telephonic survey and a field survey. Two different field surveys (one for HAR and one for CBRAS) were designed and implemented to obtain feedback on these technologies. The field survey for CBRAS is unique and has never been done before for this purpose. A sample size of 1000 HAR surveys was collected through the CATI phone survey. Field surveys were collected at five locations across the state, including central, southeast, and southwest regions of Florida. The HAR field survey sample size was 1610 and the CBRAS field survey sample size was 613. All field surveys were conducted by UCF students at each of the five locations, over a 13-week data collection period. The HAR messages were designed to alert drivers of any adverse roadway traffic or weather conditions. The CBRAS is limited to truck drivers with the closed system radio pre-installed in their vehicles. However, truck drivers were also asked some questions on HAR if they do not use CBRAS. Basic statistical analysis was used to determine a number of performance indicators which include system's use and awareness, usability of provided information, route diversion, and travelers' demographics. In addition, the two HAR phone and field samples were combined together and examined using a decision tree model. Target questions were selected from the survey to build the tree network. The tree model aimed at identifying trends between categorical differences of travelers with respect to specific questions. Understanding travelers' satisfaction with HAR is critical to knowing its benefits. The ending results indicated that both basic statistical analysis and the decision tree model are in agreement. A comparison between HAR phone and field surveys indicates the following. Travelers interviewed for the HAR field survey were more aware of the HAR than travelers surveyed by phone. A small portion of the surveyed samples used HAR (22% and this was consistent between the phone and the field surveys). Also, 80% or more were satisfied with HAR for both phone and field samples and the majority (85% or more) supported its continuation as an indication of willingness to use it in the future, especially in emergency conditions. In terms of the types of messages they want to hear from HAR, traffic congestion was the most common. Dynamic Message Signs (DMS) were the most preferred source of travel information and were the alternative for HAR, if HAR gets terminated. This was followed by smartphone applications which received twice as much support from field surveyed travelers (28%) when compared to phone surveyed travelers (15%). The CATI Phone Survey was biased towards elderly people (60% of the sample) and mainly females (58%) that use the FTE roadway system. Users satisfied with the system are those who only use these roadways once per week or less. The survey ultimately shows that travelers rely on modern modes of obtaining traffic information than traditional ones, such as HAR. DMS, and smart phone applications are leading communication tools among all type of travelers. The HAR field survey was less biased with respect to age and gender distribution (56% were under 50 and 62% were males). Both surveys indicate that the sample is well educated (about 60% have an associate degree or higher). CBRAS serves a small segment of commercial truck drivers (only 12% out of 613 used CBRAS). However, this small segment used it heavily (84% used it sometimes, often, or always). And 92% of CBRAS users were satisfied or strongly satisfied with it. CBRAS was used mostly for route divergence, with 72% of the drivers relying on it for this purpose. Truck drivers who never used CBRAS (88% of the sample) were asked questions about HAR. Only 27% of them used HAR and 57% of these used it sometimes, often, or always with 72% of the truck users being satisfied with HAR compared to the 92% satisfied with CBRAS. The most common complaint about HAR by truck drivers was that it is not easy to access or understand. Based on responses of truck drivers for both HAR and CBRAS field surveys above, it seems that GPS navigation was the most preferred source of travel information (28%). In addition to the basic statistics, a decision tree model, using SAS Enterprise Miner was performed. The statistical analysis results indicated satisfaction of travelers. The decision tree model was used to predict and profile responses to all answered questions that each survey shared. Training data was included in the model and the model was able to leverage the questions. Results of the decision tree model predicted high user satisfaction rates. Analyses of the three implemented surveys show that HAR and CBRAS technologies are not used by a large proportion of travelers, but their users are typically satisfied with these technologies. A small portion of the surveyed sample of truck drivers uses CBRAS but they use it heavily and were very satisfied with it. The travelers' satisfaction level with HAR was high. The HAR and CBRAS systems are in the middle of a heated competition lead by digital communication, it may be a sign of the time to create HAR/CBRAS smart phone applications for the longevity of these traditional technologies.