This dissertation acknowledges the fact that research regarding deaf student writers at the post-secondary level is practically void. To initiate an avenue of research that is meant to foreground future research regarding support services for deaf college student writers, I set out to find how college institutions are serving deaf student writers through academic support services by designing three institutional case studies. The first goal of this project was to synthesize research on the topic of deaf college student writers by organizing existing literature relevant to deaf education and language acquisition. The second goal of this project was to generate descriptive portraits of three institutions by illustrating how the schools are serving deaf student writers. The synthesis of the literature, descriptive portraits of the institutions, and discussion of emergent themes revealed from my study will speak to how the fields of Deaf Studies, Disability Studies, and Rhetoric and Composition might proceed in order to make the most of support services for deaf student writers.

This study compared the acquisition and maintenance of an Augmentative and Alternative device (iPad application, LAMP), and American Sign Language when teaching a 6th-grade student with an intellectual and developmental disability (IDD) and limited functional vocal verbal speech to make a request. A single-case alternating treatment design was applied to compare the acquisition rate between the two strategies. The system of least prompts was used to teach the student how to perform the request using the AAC device and ASL (American Sign Language). Results showed the student required fewer sessions to reach mastery when making a request using the AAC device. This study showed the system of least prompts paired with AAC was an effective and efficient strategy for the acquisition of a targeted communication request. This study provides additional evidence of an effective strategy that could be used when identifying a priority communication system for learners with limited functional speech and IDD.

Low-resource languages, including sign languages, are a challenge for machine translation research. Given the lack of parallel corpora, current researchers must be content with a small parallel corpus in a narrow domain for training a system. For this thesis, we obtained a small parallel corpus of English text and American Sign Language gloss from The Church of Jesus Christ of Latter-day Saints. We cleaned the corpus by loading it into an open-source translation memory tool, where we removed computer markup language and split the large chunks of text into sentences and phrases, creating a total of 14,247 sentence pairs. We randomly partitioned the corpus into three sections: 70% for a training set, 10% for a development set, and 20% for a test set. After downloading and installing the open-source Moses toolkit, we went through several iterations of training, translating, and evaluating the system. The final evaluation on unseen data yielded a state-of-the-art score for a low-resource language.

L'imagerie par résonance magnétique (IRM) permet maintenant d'observer différents types de tissus avec des résolutions de plus en plus fines. L'arbre vasculaire artériel et veineux est explorable et les flux peuvent y être caractérisés de façon non invasive. Le versant artériel de l'arbre vasculaire peut être obtenu par une imagerie dite par « temps de vol » et le versant veineux par une imagerie en contraste de phase. Le développement de reconstructions de cartographies de susceptibilité magnétique (QSM) permet d'améliorer le niveau de détails atteignable sur les veines en fournissant en plus la possibilité de quantifier des paramètres physiologiques comme la saturation veineuse en oxygène. La mise en place d'algorithmes et outils dédiés permet la reconstruction in-silico d'une architecture cohérente sujet-spécifique. Par ailleurs grâce à l'emploi de séquences de la dynamique telles que le contraste de phase dynamique et l'imagerie par marquage des protons artériels du sang, les débits artériels, veineux, et perfusionnels sont mesurables. L'intégralité de ces acquisitions est non invasive, donc applicable à l'intégralité des sujets passant des IRM. Sur la base de ces données anatomiques et dynamiques, un modèle complet et sujet-spécifique de l'hydrodynamique intracrânienne est proposé. Le flux sanguin et cérébro-spinal est décrit dans ce modèle par les équations bilans fondamentales de l'hydrodynamique : conservation de la masse, conservation de la quantité de mouvement. Pour tenir compte de l'adaptation du diamètre des vaisseaux aux pressions on introduit pour chaque compartiment un paramètre d'élasticité de la paroi et une équation correspondante. Grâce aux données d'imagerie IRM, les compartiments sanguins des artères aux veines, le parenchyme cérébral et le système ventriculaire sont inclus. Le modèle permet de simuler la répartition des flux et des pressions dans les différents compartiments de la vascularisation du patient ainsi que d'évaluer les effets d'occlusions localisées sur l'ensemble de l'architecture. / The magnetic resonance imaging (MRI) allows the observation of various kind of tissues with always increasing resolution. The arterial and venous vascular trees can be explored, and the flows can be characterized in a noninvasive way. As an example, the arterial part of the tree can be obtained using so-called “Time Of Flight” MRI, and the venous part with phase contrast techniques. The development of quantitative susceptibility maps (QSM) improves the level of details achievable regarding veins; furthermore, it provides a new way to estimate physiological parameters such as venous saturation in oxygen. Eventually the implementation of dedicated algorithms and tools allows the in-silico reconstruction of a subject-specific coherent architecture. Moreover, due to the use of dynamic imaging sequences such as the dynamic phase contrast imaging and the arterial spin labeling, the arterial, venous and cerebral blood flow are measurable. All of these sequences are noninvasive and so usable on every subjects. Based on these anatomical and dynamics data, a full subject-specific model of the brain hydrodynamics is proposed here. The blood and cerebrospinal flow are described using basic balance equations of the hydrodynamics: continuity and momentum. To take into account of the adaptation of vessel diameter to the pressure, a wall elasticity parameter is added for each compartment together with the corresponding equation. Thanks to the MRI data, all the blood compartments, from arteries to vein, the cerebral parenchyma and the ventricular system are included. The model is able to simulate the flow and pressure repartition in all compartments of the subjects as well as show the impact of a located occlusion on the whole architecture.

Les modalités d'imagerie fonctionnelle et de perfusion sont étroitement liées car les deux mesurent, directement ou indirectement, le débit sanguin cérébral. D’une part, en utilisant le contraste BOLD (Blood-Oxygen-Level-Dependent), l'imagerie fonctionnelle par résonance magnétique (IRMf) exploite les propriétés magnétiques du sang (oxy et désoxyhémoglobine) pour y mesurer les changements locaux de concentration en oxygène: ce couplage neurovasculaire permet de déduire le fonctionnement du cerveau à partir des images IRMf. D’autre part, l'IRM de perfusion reflète le fonctionnement du système vasculaire cérébral en mesurant directement le débit sanguin cérébral. En particulier, l’IRM du marquage de l’eau artérielle (ASL) n'a pas besoin d'agents de contraste: le traceur est remplacé par des spins de protons endogènes d'eau. Habituellement l’ASL est utilisée pour mesurer la perfusion basale au repos. Toutefois, ces dernières années, il a également été utilisé comme une modalité d'imagerie fonctionnelle (comme fASL) en mesurant les variations de perfusion cérébrale induites par la réalisation de tâches cognitives. Contrairement à l'IRMf standard basée sur le contraste BOLD, les résultats sont quantitatifs, ce qui rend ce type de données intéressantes pour son utilisation dans la recherche clinique.Cette thèse porte sur l’étude de la modalité fASL et sur le développement de nouvelles méthodes pour l'analyser. Comme précédemment réalisé pour les données BOLD, un cadre bayésien est développé pour l'analyse des données fASL. Il fournit un moyen de modéliser les valeurs d'activation et les fonctions de réponse hémodynamique et de perfusion en tant que variables probabilistes dans l’approche de Détection-Estimation Conjointe. Les modèles bayésiens utilisent une connaissance a priori pour l'estimation des paramètres inconnus à travers la spécification de distributions de probabilité. Dans ce travail, nous exploitons cette fonction pour incorporer au modèle des informations physiologiques, afin de rendre l'estimation plus robuste. En particulier, nous utilisons des modèles physiologiques basés sur le modèle de ballon pour obtenir un lien entre les réponses hémodynamiques et de perfusion, puis nous utilisons ce lien dans une distribution a priori pour régulariser l'estimation des réponses. En utilisant information physiologique a priori, une solution de type Markov Chain Monte Carlo (MCMC) a été proposée pour l'estimation des quantités contenues dans le signal IRMf. Étant donné que le coût de calcul de cet algorithme est très élevé, nous reformulons le problème pour utiliser une approche variationnelle (VEM) qui fournit un algorithme beaucoup plus rapide avec des résultats similaires. Dans ce cadre, l'introduction d'information a priori et de contraintes est également plus simple.Ces méthodes ont été évaluées sur deux ensembles de données différentes en utilisant des paradigmes événementiels et du bloc, pour des tâches cognitives très simples. Nous montrons les bonnes performances des méthodes proposées par rapport aux méthodes standards, au niveau des sujets et du groupe. Les résultats expérimentaux montrent que les probabilités a priori physiologiques améliorent l'estimation d'une fonction de réponse de perfusion. Ces résultats démontrent également que le contraste BOLD a une meilleure sensibilité pour la détection de l'activité cérébrale évoquée que fASL, bien que la fASL donne une activation plus localisée, ce qui est conforme à la littérature existante. A partir de ces résultats, nous discutons l'impact de la modélisation de la corrélation spatiale, ainsi que l'impact de l'estimation des réponses temporelles.Ce travail propose de nouvelles contributions méthodologiques pour l'étude de la fASL, et les met en perspective avec les techniques existantes. Ainsi, nous proposons de nouveaux outils pour la communauté neuroscientifique, mis en œuvre en python dans le package PyHRF, pour étudier et comprendre le fonctionnement du cerveau. / Functional and perfusion imaging modalities are closely related since they both measure, directly or indirectly, blood flow in the brain. Functional Magnetic Resonance Imaging (fMRI) using the blood oxygen level dependent (BOLD) contrast exploits the magnetic properties of blood (oxy- and deoxyhemoglobin) to measure local changes in blood oxygen concentration in the brain. The neurovascular coupling allows us to infer brain function from fMRI images. Perfusion MRI images the cerebral vascular system by directly measuring blood flow. In particular, Arterial Spin Labeling (ASL) does not need contrast agents; it uses spins of endogenous water protons as a tracer instead. Usually ASL is used to probe the basal perfusion at rest. However, in the recent years, it has also been used as a functional imaging modality (as fASL) by tracking task-related perfusion changes. In contrast to the standard BOLD fMRI, results are quantitative, making this type of data attractive for use in clinical research.This thesis focuses on the investigation of the fASL modality and the development of new methods to analyze it. As previously done for BOLD data, a Bayesian framework is proposed for the analysis of fASL data. It provides a way of modeling activation values and both hemodynamic and perfusion response functions as probabilistic variables in the so-called joint detection estimation (JDE) framework. Bayesian models use a priori knowledge in the estimation of unknown parameters through the specification of probability distributions. In this work, we exploit this feature to incorporate physiological information to make the estimation more robust. In particular, we use physiological models based on the balloon model to derive a link between hemodynamic and perfusion responses and we turn this link into a prior distribution to regularize the estimation of the responses. A Markov Chain Monte Carlo solution with prior physiological knowledge has been first proposed for the estimation of the quantities contained in the fMRI signal. Since the computational cost of this algorithm is very high, we then reformulate the problem to use a variational expectation maximization approach that provides a much faster algorithm with similar results. The use of priors and constraints in this setting is also more straightforward.These methods have been evaluated on two different datasets using event-related and block designs with very simple experimental tasks. We show the performance of the methods investigated in comparison to standard methods at the subject and group levels. Experimental results show the utility of using physiological priors for improving the recovery of a perfusion response function. They also demonstrate that BOLD fMRI achieves better sensitivity to detect evoked brain activity as compared to fASL although fASL gives a more localized activation, which is in line with the existing literature. From the results, we discuss the impact of the modelling of spatial correlation, as well as the impact of the estimation of temporal responses.This work proposes new methodological contributions in the study of a relatively new fMRI modality that is functional ASL, and puts it into perspective with the existing techniques. Thus, we provide new tools for the neuroscientific community to study and understand brain function. These tools have been implemented in python in the PyHRF package.

No description available.

Denna litteraturstudie syftar till att via forskning söka svar på huruvida metoden att skriva sig till läsning (ASL) är mer effektiv än andra metoder för läs- och skrivinlärning i grundskolans tidiga år. Studiens frågeställningar har besvarats genom en systematisk litteraturstudie där vetenskaplig litteratur har granskats och analyserats. För att söka vetenskapliga artiklar har databaserna Google Scholar och ERIC (Education Resources Information Center) använts. Personer som förespråkar metoden har även kontaktats och tipsat om den forskning de känner till. Det framkom genom litteratursökningen att det finns begränsat med forskning kring ASL och den forskning som finns efterfrågar ytterligare forskning. Vissa forskningsresultat pekar på att ASL kan vara effektivt för elevers läs- och skrivinlärning men resultatet av denna litteraturstudie visar hur komplext det kan vara att undersöka effektiviteten av en metod då det kan vara svårt att finna en kontrollgrupp som har samma förutsättningar som testgruppen. ASL blir allt mer populärt i svenska skolor och många som använder metoden hyllar den, men det är inte tydligt om det är det datorbaserade skrivandet i sig eller andra faktorer som bidrar till metodens påstådda framgångar. Under rubriken diskussion diskuteras även möjligheterna och begränsningarna kring forskning av läs- och skrivinlärningsmetoder.

Cystic fibrosis (CF) is a lethal genetic disease that affects 30,000 people in the United States alone. While the disease affects organs throughout the body, it is the lung disease that is the primary cause of morbidity and mortality for people with the disease. CF lung disease is characterized by thick and sticky mucus that obstructs the airways, acute and chronic bacterial infections, and chronic inflammation and remodeling. Thanks to the creation of the CF pig, it is now possible to study the manifestations of CF lung disease at birth. The CF pig develops spontaneous lung disease, similar to that found in humans with CF, making it the ideal model for our studies. One of the critical findings that revealed in studies of the CF pig is that airway surface liquid (ASL) bactericidal activity is impaired in CF at birth, and this activity is pH dependent. Because infants and children with CF tend to suffer greater morbidity from respiratory viruses than non-CF infants and children, we sought to determine if ASL has antiviral activity and if that activity is reduced in newborn CF pigs. We found that pre-incubating either tracheal or nasal ASL from wild-type pigs reduced the infectivity of various recombinant viruses expressing an eGFP or GFP reporter gene. Those viruses include Sendai virus (SeV-eGFP), respiratory syncytial virus (RSV-GFP), the PR8 strain of influenza virus A (PR8-eGFP), and adenovirus (Ad-eGFP), indicating ASL has broad-spectrum antiviral activity. Nasal secretions from newborn CF pigs had strikingly reduced antiviral activity against SeV-eGFP and Ad-eGFP compared to nasal secretions from WT littermates. Unlike what was observed for ASL antibacterial activity, nasal secretion antiviral activity was not affected by pH, nor was it affected by bicarbonate concentration, one of the molecules that drives pH in the airways. However, when we mixed CF and WT nasal secretions at different ratios, we found the antiviral activity to follow a linear trend, with antiviral activity increasing as the percentage of WT nasal secretions increased. This suggests that one or more components of nasal secretions are found less abundantly in CF nasal secretions compared to WT nasal secretions, leading to reduced antiviral activity in CF. The CF pig has facilitated a much greater understanding of the early stages of CF lung disease. This model will allow us to determine what antiviral components are lacking in the CF airways and why they are reduced in CF.

Deaf and hard of hearing (d/hh) children often acquire an L1 after age 3, thus are arguably more diverse than that of the general bilingual population. A unique problem therefore exists among d/hh late language learners—they often do not have an L1 to later develop an L2. This study investigated the impact of an English writing intervention (Strategic and Interactive Writing Instruction, SIWI) that incorporates support for the development of American Sign Language in an effort to illustrate the necessity of explicitly addressing the proposed interdependence of language learning. The research involved providing 23 upper elementary and middle school d/hh students with SIWI. SIWI has been shown to have a significant impact on student outcomes in language and literacy. The study was conducted in five classrooms—one fourth, two fifth, and two sixth grade classrooms—over a twelve-week period at a state residential school for the deaf. This allowed for two weeks of pre-test, mid-test and post-test administration, five weeks of regular instruction, and five weeks of intervention. The students received SIWI for four forty-five minute sessions and one thirty-minute session each week for a total of five weeks. The intervention replaced their regular 45 minutes of writing instruction. In order to measure expressive language growth in ASL, language samples for each student participant were collected. These samples were analyzed to chart expressive language growth during the time period with no SIWI intervention and while engaged in SIWI by reviewing them for students’ mean length of utterance (MLU), use of unintelligible utterances, and specific grammatical features of ASL, and individually for patterns of ASL expressive language growth. Repeated measures ANOVAs (within and between subjects) conducted for students’ MLU and unintelligible utterances revealed statistically significant growth after five weeks of SIWI. This study demonstrates the reciprocity of language learning. The foregrounding of written English supported the development of a more nuanced understanding of the use and features of ASL.

Syftet med studien var att undersöka hur rektor, speciallärare och lärare förhåller sig till metoden ASL på tre olika skolor. Vi har använt oss av intervjuer med ljudinspelningar och observationer för att samla in data. Som teori och stöd för analysen har didaktikens fyra grundfrågor vad, hur, varför och vem använts. En av forskningsfrågorna var även att undersöka vad informanterna anser är för- och nackdelarna med metoden ASL. Flera informanter anser att en fördel med metoden är att det är lätt att individanpassa, samt att en nackdel är om tekniken inte fungerar.