Дисертації з теми "Artificial magnetic resonance"

Abstract Eddy currents and susceptibility differences are the most important sources that interfere with the quality of MR images in the presence of an artificial macroscopic object in the volume to be imaged. In this study, both of these factors have been examined. The findings show that the RF field is the most important cause of induced eddy currents when gradients with relatively slow slew rates are used. The induced eddy currents amplify or dampen the RF field with the result that the flip angle changes. At the proximal end in the vicinity of the hip prosthesis surface, there have been areas where the flip angle is nearly threefold compared to the reference flip angle. Areas with decreased flip angles have also been found near the surface of the prosthesis top. The incompleteness of the image due to eddy currents manifests as signal loss areas. Two different methods based on MRI were developed to estimate the susceptibility of a cylindrical object. One of them is based on geometrical distortions in SE magnitude images, while the other takes advantage of phase differences in GRE phase images. The estimate value of the Profile™ test hip prosthesis is χ = (170 ± 13) 10-6. A remapping method was selected to correct susceptibility image distortions. Correction was accomplished with pixel shifts in the frequency domain. The magnetic field distortions were measured using GRE phase images. The method was tested by simulations and by imaging a hip prosthesis in a water tank and in a human pelvis. The main limitations of the method described here are the loss of a single-valued correction map with higher susceptibility differences and the problems with phase unwrapping in phase images. Modulation transfer functions (MTF) were exploited to assess the effect of correction procedure. The corrected image of a prosthesis in a human hip after total hip arthroplasty appears to be equally sharp or slightly sharper than the corresponding original images. The computer programs written for this study are presented in an appendix.

In this study a hybrid Pore Network (PN) model that simulates two-phase (water-oil) drainage and imbibition mechanisms is developed. The developed model produces Nuclear Magnetic Resonance (NMR) T2 relaxation times using correlations available in the literature. The developed PN was calibrated using experimental relative permeability data obtained for Berea Sandstone, Kuzey Marmara Limestone, Yenikö
y Dolostone and Dolomitic Limestone core plugs. Pore network body and throat parameters were obtained from serial computerized tomography scans and thin section images. It was observed that pore body and throat sizes were not statistically correlated. It was also observed that the developed PN model can be used to model different displacement mechanisms. By using the synthetic data obtained from PN model, an Artificial Neural Network (ANN) model was developed and tested. It has been observed that the developed ANN tool can be used to estimate oil &ndash
water relative permeability data very well (with less than 0.05 mean square error) given a T2 signal. It was finally concluded that the developed tools can be used to obtain multiphase flow functions directly from an NMR well log such as Combinable Magnetic Resonance (CMR).

A better understanding of tumour heterogeneity is central for accurate diagnosis, targeted therapy and personalised treatment of glioblastoma patients. This thesis aims to investigate whether pre-operative multi-parametric magnetic resonance imaging (MRI) can provide a useful tool for evaluating inter-tumoural and intra-tumoural heterogeneity of glioblastoma. For this purpose, we explored: 1) the utilities of habitat imaging in combining multi-parametric MRI for identifying invasive sub-regions (I & II); 2) the significance of integrating multi-parametric MRI, and extracting modality inter-dependence for patient stratification (III & IV); 3) the value of advanced physiological MRI and radiomics approach in predicting epigenetic phenotypes (V). The following observations were made: I. Using a joint histogram analysis method, habitats with different diffusivity patterns were identified. A non-enhancing sub-region with decreased isotropic diffusion and increased anisotropic diffusion was associated with progression-free survival (PFS, hazard ratio [HR] = 1.08, P < 0.001) and overall survival (OS, HR = 1.36, P < 0.001) in multivariate models. II. Using a thresholding method, two low perfusion compartments were identified, which displayed hypoxic and pro-inflammatory microenvironment. Higher lactate in the low perfusion compartment with restricted diffusion was associated with a worse survival (PFS: HR = 2.995, P = 0.047; OS: HR = 4.974, P = 0.005). III. Using an unsupervised multi-view feature selection and late integration method, two patient subgroups were identified, which demonstrated distinct OS (P = 0.007) and PFS (P < 0.001). Features selected by this approach showed significantly incremental prognostic value for 12-month OS (P = 0.049) and PFS (P = 0.022) than clinical factors. IV. Using a method of unsupervised clustering via copula transform and discrete feature extraction, three patient subgroups were identified. The subtype demonstrating high inter-dependency of diffusion and perfusion displayed higher lactate than the other two subtypes (P = 0.016 and P = 0.044, respectively). Both subtypes of low and high inter-dependency showed worse PFS compared to the intermediate subtype (P = 0.046 and P = 0.009, respectively). V. Using a radiomics approach, advanced physiological images showed better performance than structural images for predicting O6-methylguanine-DNA methyltransferase (MGMT) methylation status. For predicting 12-month PFS, the model of radiomic features and clinical factors outperformed the model of MGMT methylation and clinical factors (P = 0.010). In summary, pre-operative multi-parametric MRI shows potential for the non-invasive evaluation of glioblastoma heterogeneity, which could provide crucial information for patient care.

Includes vita.
Title from PDF of title page.
Document formatted into pages; contains 222 pages.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: This dissertation presents a knowledge-guided expert system that is capable of applying routinesfor multispectral analysis, (un)supervised clustering, and basic image processing to automatically detect and segment brain tissue abnormalities, and then label glioblastoma-multiforme brain tumors in magnetic resonance volumes of the human brain. The magnetic resonance images used here consist of three feature images (T1-weighted, proton density, T2-weighted) and the system is designed to be independent of a particular scanning protocol. Separate, but contiguous 2D slices in the transaxial plane form a brain volume. This allows complete tumor volumes to be measured and if repeat scans are taken over time, the system may be used to monitor tumor response to past treatments and aid in the planning of future treatment. Furthermore, once processing begins, the system is completely unsupervised, thus avoiding the problems of human variability found in supervised segmentation efforts.Each slice is initially segmented by an unsupervised fuzzy c-means algorithm. The segmented image, along with its respective cluster centers, is then analyzed by a rule-based expert system which iteratively locates tissues of interest based on the hierarchy of cluster centers in feature space. Model-based recognition techniques analyze tissues of interest by searching for expected characteristics and comparing those found with previously defined qualitative models. Normal/abnormal classification is performed through a default reasoning method: if a significant model deviation is found, the slice is considered abnormal. Otherwise, the slice is considered normal. Tumor segmentation in abnormal slices begins with multispectral histogram analysis and thresholding to separate suspected tumor from the rest of the intra-cranial region. The tumor is then refined with a variant of seed growing, followed by spatial component analysis and a final thresholding step to remove non-tumor pixels.The knowledge used in this system was extracted from general principles of magnetic resonance imaging, the distributions of individual voxels and cluster centers in feature space, and anatomical information. Knowledge is used both for single slice processing and information propagation between slices. A standard rule-based expert system shell (CLIPS) was modified to include the multispectral analysis, clustering, and image processing tools.A total of sixty-three volume data sets from eight patients and seventeen volunteers (four with and thirteen without gadolinium enhancement) were acquired from a single magnetic resonance imaging system with slightly varying scanning protocols were available for processing. All volumes were processed for normal/abnormal classification. Tumor segmentation was performed on the abnormal slices and the results were compared with a radiologist-labeled ground truth' tumor volume and tumor segmentations created by applying supervised k-nearest neighbors, a partially supervised variant of the fuzzy c-means clustering algorithm, and a commercially available seed growing package. The results of the developed automatic system generally correspond well to ground truth, both on a per slice basis and more importantly in tracking total tumor volume during treatment over time.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.

[ES] El futuro de la imagen médica está ligado a la inteligencia artificial. El análisis manual de imágenes médicas es hoy en día una tarea ardua, propensa a errores y a menudo inasequible para los humanos, que ha llamado la atención de la comunidad de Aprendizaje Automático (AA). La Imagen por Resonancia Magnética (IRM) nos proporciona una rica variedad de representaciones de la morfología y el comportamiento de lesiones inaccesibles sin una intervención invasiva arriesgada. Sin embargo, explotar la potente pero a menudo latente información contenida en la IRM es una tarea muy complicada, que requiere técnicas de análisis computacional inteligente. Los tumores del sistema nervioso central son una de las enfermedades más críticas estudiadas a través de IRM. Específicamente, el glioblastoma representa un gran desafío, ya que, hasta la fecha, continua siendo un cáncer letal que carece de una terapia satisfactoria. Del conjunto de características que hacen del glioblastoma un tumor tan agresivo, un aspecto particular que ha sido ampliamente estudiado es su heterogeneidad vascular. La fuerte proliferación vascular del glioblastoma, así como su robusta angiogénesis han sido consideradas responsables de la alta letalidad de esta neoplasia. Esta tesis se centra en la investigación y desarrollo del método Hemodynamic Tissue Signature (HTS): un método de AA no supervisado para describir la heterogeneidad vascular de los glioblastomas mediante el análisis de perfusión por IRM. El método HTS se basa en el concepto de hábitat, que se define como una subregión de la lesión con un perfil de IRM que describe un comportamiento fisiológico concreto. El método HTS delinea cuatro hábitats en el glioblastoma: el hábitat HAT, como la región más perfundida del tumor con captación de contraste; el hábitat LAT, como la región del tumor con un perfil angiogénico más bajo; el hábitat IPE, como la región adyacente al tumor con índices de perfusión elevados; y el hábitat VPE, como el edema restante de la lesión con el perfil de perfusión más bajo. La investigación y desarrollo de este método ha originado una serie de contribuciones enmarcadas en esta tesis. Primero, para verificar la fiabilidad de los métodos de AA no supervisados en la extracción de patrones de IRM, se realizó una comparativa para la tarea de segmentación de gliomas de grado alto. Segundo, se propuso un algoritmo de AA no supervisado dentro de la familia de los Spatially Varying Finite Mixture Models. El algoritmo propone una densidad a priori basada en un Markov Random Field combinado con la función probabilística Non-Local Means, para codificar la idea de que píxeles vecinos tienden a pertenecer al mismo objeto. Tercero, se presenta el método HTS para describir la heterogeneidad vascular del glioblastoma. El método se ha aplicado a casos reales en una cohorte local de un solo centro y en una cohorte internacional de más de 180 pacientes de 7 centros europeos. Se llevó a cabo una evaluación exhaustiva del método para medir el potencial pronóstico de los hábitats HTS. Finalmente, la tecnología desarrollada en la tesis se ha integrado en la plataforma online ONCOhabitats (https://www.oncohabitats.upv.es). La plataforma ofrece dos servicios: 1) segmentación de tejidos de glioblastoma, y 2) evaluación de la heterogeneidad vascular del tumor mediante el método HTS. Los resultados de esta tesis han sido publicados en diez contribuciones científicas, incluyendo revistas y conferencias de alto impacto en las áreas de Informática Médica, Estadística y Probabilidad, Radiología y Medicina Nuclear y Aprendizaje Automático. También se emitió una patente industrial registrada en España, Europa y EEUU. Finalmente, las ideas originales concebidas en esta tesis dieron lugar a la creación de ONCOANALYTICS CDX, una empresa enmarcada en el modelo de negocio de los companion diagnostics de compuestos farmacéuticos.
[EN] The future of medical imaging is linked to Artificial Intelligence (AI). The manual analysis of medical images is nowadays an arduous, error-prone and often unaffordable task for humans, which has caught the attention of the Machine Learning (ML) community. Magnetic Resonance Imaging (MRI) provides us with a wide variety of rich representations of the morphology and behavior of lesions completely inaccessible without a risky invasive intervention. Nevertheless, harnessing the powerful but often latent information contained in MRI acquisitions is a very complicated task, which requires computational intelligent analysis techniques. Central nervous system tumors are one of the most critical diseases studied through MRI. Specifically, glioblastoma represents a major challenge, as it remains a lethal cancer that, to date, lacks a satisfactory therapy. Of the entire set of characteristics that make glioblastoma so aggressive, a particular aspect that has been widely studied is its vascular heterogeneity. The strong vascular proliferation of glioblastomas, as well as their robust angiogenesis and extensive microvasculature heterogeneity have been claimed responsible for the high lethality of the neoplasm. This thesis focuses on the research and development of the Hemodynamic Tissue Signature (HTS) method: an unsupervised ML approach to describe the vascular heterogeneity of glioblastomas by means of perfusion MRI analysis. The HTS builds on the concept of habitats. A habitat is defined as a sub-region of the lesion with a particular MRI profile describing a specific physiological behavior. The HTS method delineates four habitats within the glioblastoma: the HAT habitat, as the most perfused region of the enhancing tumor; the LAT habitat, as the region of the enhancing tumor with a lower angiogenic profile; the potentially IPE habitat, as the non-enhancing region adjacent to the tumor with elevated perfusion indexes; and the VPE habitat, as the remaining edema of the lesion with the lowest perfusion profile. The research and development of the HTS method has generated a number of contributions to this thesis. First, in order to verify that unsupervised learning methods are reliable to extract MRI patterns to describe the heterogeneity of a lesion, a comparison among several unsupervised learning methods was conducted for the task of high grade glioma segmentation. Second, a Bayesian unsupervised learning algorithm from the family of Spatially Varying Finite Mixture Models is proposed. The algorithm integrates a Markov Random Field prior density weighted by the probabilistic Non-Local Means function, to codify the idea that neighboring pixels tend to belong to the same semantic object. Third, the HTS method to describe the vascular heterogeneity of glioblastomas is presented. The HTS method has been applied to real cases, both in a local single-center cohort of patients, and in an international retrospective cohort of more than 180 patients from 7 European centers. A comprehensive evaluation of the method was conducted to measure the prognostic potential of the HTS habitats. Finally, the technology developed in this thesis has been integrated into an online open-access platform for its academic use. The ONCOhabitats platform is hosted at https://www.oncohabitats.upv.es, and provides two main services: 1) glioblastoma tissue segmentation, and 2) vascular heterogeneity assessment of glioblastomas by means of the HTS method. The results of this thesis have been published in ten scientific contributions, including top-ranked journals and conferences in the areas of Medical Informatics, Statistics and Probability, Radiology & Nuclear Medicine and Machine Learning. An industrial patent registered in Spain, Europe and EEUU was also issued. Finally, the original ideas conceived in this thesis led to the foundation of ONCOANALYTICS CDX, a company framed into the business model of companion diagnostics for pharmaceutical compounds.
[CA] El futur de la imatge mèdica està lligat a la intel·ligència artificial. L'anàlisi manual d'imatges mèdiques és hui dia una tasca àrdua, propensa a errors i sovint inassequible per als humans, que ha cridat l'atenció de la comunitat d'Aprenentatge Automàtic (AA). La Imatge per Ressonància Magnètica (IRM) ens proporciona una àmplia varietat de representacions de la morfologia i el comportament de lesions inaccessibles sense una intervenció invasiva arriscada. Tanmateix, explotar la potent però sovint latent informació continguda a les adquisicions de IRM esdevé una tasca molt complicada, que requereix tècniques d'anàlisi computacional intel·ligent. Els tumors del sistema nerviós central són una de les malalties més crítiques estudiades a través de IRM. Específicament, el glioblastoma representa un gran repte, ja que, fins hui, continua siguent un càncer letal que manca d'una teràpia satisfactòria. Del conjunt de característiques que fan del glioblastoma un tumor tan agressiu, un aspecte particular que ha sigut àmpliament estudiat és la seua heterogeneïtat vascular. La forta proliferació vascular dels glioblastomes, així com la seua robusta angiogènesi han sigut considerades responsables de l'alta letalitat d'aquesta neoplàsia. Aquesta tesi es centra en la recerca i desenvolupament del mètode Hemodynamic Tissue Signature (HTS): un mètode d'AA no supervisat per descriure l'heterogeneïtat vascular dels glioblastomas mitjançant l'anàlisi de perfusió per IRM. El mètode HTS es basa en el concepte d'hàbitat, que es defineix com una subregió de la lesió amb un perfil particular d'IRM, que descriu un comportament fisiològic concret. El mètode HTS delinea quatre hàbitats dins del glioblastoma: l'hàbitat HAT, com la regió més perfosa del tumor amb captació de contrast; l'hàbitat LAT, com la regió del tumor amb un perfil angiogènic més baix; l'hàbitat IPE, com la regió adjacent al tumor amb índexs de perfusió elevats, i l'hàbitat VPE, com l'edema restant de la lesió amb el perfil de perfusió més baix. La recerca i desenvolupament del mètode HTS ha originat una sèrie de contribucions emmarcades a aquesta tesi. Primer, per verificar la fiabilitat dels mètodes d'AA no supervisats en l'extracció de patrons d'IRM, es va realitzar una comparativa en la tasca de segmentació de gliomes de grau alt. Segon, s'ha proposat un algorisme d'AA no supervisat dintre de la família dels Spatially Varying Finite Mixture Models. L'algorisme proposa un densitat a priori basada en un Markov Random Field combinat amb la funció probabilística Non-Local Means, per a codificar la idea que els píxels veïns tendeixen a pertànyer al mateix objecte semàntic. Tercer, es presenta el mètode HTS per descriure l'heterogeneïtat vascular dels glioblastomas. El mètode HTS s'ha aplicat a casos reals en una cohort local d'un sol centre i en una cohort internacional de més de 180 pacients de 7 centres europeus. Es va dur a terme una avaluació exhaustiva del mètode per mesurar el potencial pronòstic dels hàbitats HTS. Finalment, la tecnologia desenvolupada en aquesta tesi s'ha integrat en una plataforma online ONCOhabitats (https://www.oncohabitats.upv.es). La plataforma ofereix dos serveis: 1) segmentació dels teixits del glioblastoma, i 2) avaluació de l'heterogeneïtat vascular dels glioblastomes mitjançant el mètode HTS. Els resultats d'aquesta tesi han sigut publicats en deu contribucions científiques, incloent revistes i conferències de primer nivell a les àrees d'Informàtica Mèdica, Estadística i Probabilitat, Radiologia i Medicina Nuclear i Aprenentatge Automàtic. També es va emetre una patent industrial registrada a Espanya, Europa i els EEUU. Finalment, les idees originals concebudes en aquesta tesi van donar lloc a la creació d'ONCOANALYTICS CDX, una empresa emmarcada en el model de negoci dels companion diagnostics de compostos farmacèutics.
En este sentido quiero agradecer a las diferentes instituciones y estructuras de financiación de investigación que han contribuido al desarrollo de esta tesis. En especial quiero agradecer a la Universitat Politècnica de València, donde he desarrollado toda mi carrera acadèmica y científica, así como al Ministerio de Ciencia e Innovación, al Ministerio de Economía y Competitividad, a la Comisión Europea, al EIT Health Programme y a la fundación Caixa Impulse
Juan Albarracín, J. (2020). Unsupervised learning for vascular heterogeneity assessment of glioblastoma based on magnetic resonance imaging: The Hemodynamic Tissue Signature [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/149560
TESIS

Garantir a qualidade da carne bovina produzida no Brasil tem sido uma preocupação dos produtores, pois contribui para aumentar a exportação e o consumo interno do produto. Por isso, tem-se pesquisado novos métodos que analisam e garantam a qualidade da carne, de forma rápida, eficiente e não destrutiva. A ressonância magnética nuclear (RMN) tem se destacado como uma das técnicas de controle de qualidade de carne. Neste trabalho as redes neurais artificiais estão sendo utilizadas para o reconhecimento de padrões dos dados de ressonância magnética nuclear oriundos de carne bovina. Mais especificamente, os respectivos dados têm sido utilizados por uma rede perceptron multicamadas para a extração de características da carne bovina, possibilitando a classificação do grupo genético e do sexo dos animais a partir de uma amostra da referida carne. Os resultados dos experimentos são também apresentados para ilustrar o desempenho da abordagem proposta.
Guaranteeing the quality of the bovine meat produced in Brazil has been a concern of the producers because it contributes to increase the export and the domestic consumption of the product. Therefore, new methods have been researched that analyze and guarantee the quality of the meat in a fast, efficient and non destructive way. Nuclear magnetic resonance (NMR) has been highlighted as one of the techniques of meat quality control. In this work study artificial neural networks are being used for pattern recognition from data obtained by the resonance equipment, originating from bovine meat. More specifically, the respective data have been used by a multilayer perceptron network for extraction of bovine meat characteristics, making possible the classification of both genetic group and animal sex starting from a single meat sample. Several results of experimental tests are also presented to illustrate the performance of the proposed approach.

A reliable evaluation system is essential for every automatic process. While techniques for automatic segmentation of images have been extensively researched in recent years, evaluation of the same has not received an equal amount of attention. Amra Medical AB has developed a system for automatic segmentation of magnetic resonance (MR) images of human bodies using an atlas-based approach. Through their software, Amra is able to derive body composition measurements, such as muscle and fat volumes, from the segmented MR images. As of now, the automatic segmentations are quality controlled by clinical experts to ensure their correctness. This thesis investigates the possibilities to leverage predictive modelling to reduce the need for a manual quality control (QC) step in an otherwise automatic process. Two different regression approaches have been implemented as a part of this study: body composition measurement prediction (BCMP) and manual correction prediction (MCP). BCMP aims at predicting the derived body composition measurements and comparing the predictions to actual measurements. The theory is that large deviations between the predictions and the measurements signify an erroneously segmented sample. MCP instead tries to directly predict the amount of manual correction needed for each sample. Several regression models have been implemented and evaluated for the two approaches. Comparison of the regression models shows that local linear regression (LLR) is the most performant model for both BCMP and MCP. The results show that the inaccuracies in the BCMP-models, in practice, renders this approach useless. MCP proved to be a far more viable approach; using MCP together with LLR achieves a high true positive rate with a reasonably low false positive rate for several body composition measurements. These results suggest that the type of system developed in this thesis has the potential to reduce the need for manual inspections of the automatic segmentation masks.

Introduktion: Uppskattningsvis 40% av alla magnetresonanstomografiska (MRT) undersökningar som görs i Europa och USA utförs med gadoliniumbaserade kontrastmedel. Under det senaste decenniet har det i flera studier uppmärksammats att ansamling av gadolinium sker i olika strukturer i hjärnan. Patienter med multipel skleros följs regelbundet upp med MRT undersökningar och MRT med kontrastförstärkning är den vanligaste metoden för att urskilja nytillkomna patologiska förändringar. Utveckling inom teknologi och metoder inom artificiell intelligens har visat att det finns anledning att kartlägga om röntgensjuksköterskans arbete med undersökningar och läkemedel som administreras till patienter kan förändras så att det förebygger ansamling av gadolinium. Syfte: Syftet med denna litteraturöversikt var att kartlägga ansamlingen av gadoliniumkontrastmedel hos patienter med multipel skleros och hur artificiell intelligens kan tillämpas vid MRT för att minska användning av gadoliniumkontrast. Metod: Allmän litteraturöversikt där vetenskapliga artiklar av kvantitativ studiedesign har sökts fram genom databaserna CINAHL och PubMed. Resultat: Både makrocykliska och linjära gadoliniumbaserade kontrastmedel ansamlas i de basala ganglierna. Genom tillämpning av AI och CAD går det att framställa bilder med fullgod bildkvalitet och samtidigt reducera mängden kontrastmedel som administreras till patienten. Slutsats: Det behövs mer forskning om gadoliniumansamling för att nya rutiner och metoder ska kunna implementeras. Ansamling av gadolinium visar att det finns skäl att fortsätta utveckla nya metoder för uppföljning av sjukdomsförloppet hos MS-patienter. När det gäller AI inom medicinsk bilddiagnostik och magnetresonanstomografi finns många utvecklingsmöjligheter som kan bidra till minskning av gadoliniumbaserad kontrast i framtiden. Fortsatt forskning inom deep learning och CAD kan i framtiden utvecklas så att röntgensjuksköterskan får en mer självbestämmande funktion i bildframställning vid MRT, men även ett mer självständigt arbete i hanteringen av farmaka. Dessutom kan denna utveckling bidra till att röntgensjuksköterskans multidiciplinära samverkan med radiologer stärks och bidrar till en positiv utveckling med kortare granskningstider, bättre hantering av patienter, optimerade undersökningar, minskning av undersökningstider och kortare vårdköer.
Introduction: Approximately 40% of all magnetic resonance imaging (MRI) scans performed in Europe and the United States are performed with gadolinium based contrast agents. Over the past decade, several studies have shown a gadolinium deposition in various structures in the brain. Patients with multiple sclerosis are regularly followed up with MRI with contrast enhancement is the most common method for distinguishing new pathological changes. Developments in technology and methods in artificial intelligence have shown that there is reason to map out whether the radiographers work with examinations and drugs administered to patients can be changed so that the accumulation of gadolinium is prevented. Aim: The purpose of this literature review was to examine the accumulation of gadolinium contrast agents in patients with multiple sclerosis of gadolinium contrast agents in patients with multiple sclerosis and how artificial intelligence can be applied in MRI to reduce the use of gadolinium based contrast agents. Methods: General literature review where scientific articles of a quantitative nature have been searched through the databases CINAHL and PubMed. Results: Both macrocyclic and linear gadolinium based contrast agents are retained in the basal ganglia. With artificial intelligence and CAD, it is possible to obtain data with good quality and at the same time reduce the amount of gadolinium based contrasts to patients. Conclusions: More research on gadolinium accumulation is needed for new routines and methods to be implemented. Accumulation of gadolinium shows that there is reason to continue to develop new methods for monitoring the course of the disease in MS patients. Concerning AI in medical imaging and magnetic resonance imaging, there are many development opportunities that can contribute to the reduction of gadolinium contrast in the future. Continued research in deep learning and CAD can be developed in the future so that the X-ray nurse has a more self-determining function in image production in MRI, but also a more independent work in the management of pharmacies. In addition, this development can contribute to the X - ray nurse's multidisciplinary collaboration with radiologists is strengthened and contributes to a positive development in shorter examination times, better management of patients, optimized examinations, reduction of examination times and shorter care queues.

Thesis (Ph. D.)--University of California, San Diego, 2008.
Title from first page of PDF file (viewed Feb. 6, 2009). Available via ProQuest Digital Dissertations. Vita. Includes bibliographical references (p. 109-115).

O disco intervertebral é uma estrutura cuja função é receber, amortecer e distribuir o impacto das cargas impostas sobre a coluna vertebral. O aumento da idade e a postura adotada pelo indivíduo podem levar à degeneração do disco intervertebral. Atualmente, a Ressonância Magnética (RM) é considerada o melhor e mais sensível método não invasivo de avaliação por imagem do disco intervertebral. Neste trabalho foram desenvolvidos métodos quantitativos computadorizados para auxílio ao diagnóstico da degeneração do disco intervertebral em imagens de ressonância magnética ponderadas em T2 da coluna lombar, de acordo com a escala de Pfirrmann, uma escala semi-quantitativa, com cinco graus de degeneração. Os algoritmos computacionais foram testados em um conjunto de dados que consiste de imagens de 300 discos, obtidos de 102 indivíduos, com diferentes graus de degeneração. Máscaras binárias de discos segmentados manualmente foram utilizadas para calcular seus centroides, visando criar um ponto de referência para possibilitar a extração de atributos. Uma análise de textura foi realizada utilizando a abordagem proposta por Haralick. Para caracterização de forma, também foram calculados os momentos invariantes definidos por Hu e os momentos centrais para cada disco. A classificação do grau de degeneração foi realizada utilizando uma rede neural artificial e o conjunto de atributos extraídos de cada disco. Uma taxa média de acerto na classificação de 87%, com erro padrão de 6,59% e uma área média sob a curva ROC (Receiver Operating Characteristic) de 0,92 indicam o potencial de aplicação dos algoritmos desenvolvidos como ferramenta de apoio ao diagnóstico da degeneração do disco intervertebral.
The intervertebral disc is a structure whose function is to receive, absorb and transmit the impact loads imposed on the spine. Increasing age and the posture adopted by the individual can lead to degeneration of the intervertebral disc. Currently, Magnetic Resonance Imaging (MRI) is considered the best and most sensitive noninvasive method to imaging evaluation of the intervertebral disc. In this work were developed methods for quantitative computer-aided diagnosis of the intervertebral disc degeneration in MRI T2 weighted images of the lumbar column according to Pfirrmann scale, a semi-quantitative scale with five degrees of degeneration. The algorithms were tested on a dataset of 300 images obtained from 102 subjects with varying degrees of degeneration. Binary masks manually segmented of the discs were used to calculate their centroids, to create a reference point to enable extraction of attributes. A texture analysis was performed using the approach proposed by Haralick. For the shape characterization, invariant moments defined by Hu and central moments were also calculated for each disc. The rating of the degree of degeneration was performed using an artificial neural network and the set of extracted attributes of each disk. An average rate of correct classification of 87%, with standard error 6.59% and an average area under the ROC curve (Receiver Operating Characteristic) of 0.92 indicates the potential application of the algorithms developed as a diagnostic support tool to the degeneration of the intervertebral disc.

Doutoramento em Química
The main objective of this work was to monitor a set of physical-chemical properties of heavy oil procedural streams through nuclear magnetic resonance spectroscopy, in order to propose an analysis procedure and online data processing for process control. Different statistical methods which allow to relate the results obtained by nuclear magnetic resonance spectroscopy with the results obtained by the conventional standard methods during the characterization of the different streams, have been implemented in order to develop models for predicting these same properties. The real-time knowledge of these physical-chemical properties of petroleum fractions is very important for enhancing refinery operations, ensuring technically, economically and environmentally proper refinery operations. The first part of this work involved the determination of many physical-chemical properties, at Matosinhos refinery, by following some standard methods important to evaluate and characterize light vacuum gas oil, heavy vacuum gas oil and fuel oil fractions. Kinematic viscosity, density, sulfur content, flash point, carbon residue, P-value and atmospheric and vacuum distillations were the properties analysed. Besides the analysis by using the standard methods, the same samples were analysed by nuclear magnetic resonance spectroscopy. The second part of this work was related to the application of multivariate statistical methods, which correlate the physical-chemical properties with the quantitative information acquired by nuclear magnetic resonance spectroscopy. Several methods were applied, including principal component analysis, principal component regression, partial least squares and artificial neural networks. Principal component analysis was used to reduce the number of predictive variables and to transform them into new variables, the principal components. These principal components were used as inputs of the principal component regression and artificial neural networks models. For the partial least squares model, the original data was used as input. Taking into account the performance of the develop models, by analysing selected statistical performance indexes, it was possible to conclude that principal component regression lead to worse performances. When applying the partial least squares and artificial neural networks models better results were achieved. However, it was with the artificial neural networks model that better predictions were obtained for almost of the properties analysed. With reference to the results obtained, it was possible to conclude that nuclear magnetic resonance spectroscopy combined with multivariate statistical methods can be used to predict physical-chemical properties of petroleum fractions. It has been shown that this technique can be considered a potential alternative to the conventional standard methods having obtained very promising results.
O principal objetivo deste trabalho foi monitorizar um conjunto de propriedades físico-químicas de correntes processuais pesadas através da espectroscopia de ressonância magnética nuclear, com o intuito de propor um procedimento de análise e processamento de dados em linha para o controlo processual. Vários métodos estatísticos que permitiram relacionar os resultados obtidos por espectroscopia de ressonância magnética nuclear com os resultados obtidos pelos métodos convencionais, aquando da caracterização das diferentes correntes, foram implementados a fim de desenvolver modelos de previsão dessas mesmas propriedades. O conhecimento em tempo real das propriedades físico-químicas dos derivados de petróleo é essencial para otimizar as operações de refinação, garantindo assim operações técnica, económica e ambientalmente adequadas. A primeira parte deste trabalho envolveu a realização de um elevado número de experiências, efetuadas na refinaria de Matosinhos, seguindo métodos convencionais normalizados, importantes para avaliar e caracterizar as correntes de gasóleo de vácuo leve, gasóleo de vácuo pesado e fuel óleo. As propriedades analisadas foram a massa volúmica, viscosidade cinemática, teor em enxofre, ponto de inflamação, resíduo carbonoso, valor P e a destilação atmosférica e de vácuo. Para além da determinação de todas estas propriedades físico-químicas, usando os métodos convencionais, as mesmas amostras foram analisadas por espectroscopia de ressonância magnética nuclear. A segunda parte deste trabalho esteve relacionada com a aplicação de métodos estatísticos multivariáveis que relacionam as propriedades físico-químicas com a informação quantitativa adquirida por espectroscopia de ressonância magnética nuclear. Vários métodos foram aplicados, incluindo a análise por componentes principais, a regressão múltipla por componentes principais, as regressões múltiplas parciais e as redes neuronais artificiais. A análise de componentes principais foi utilizada para reduzir o número de variáveis preditivas e transformá-las em novas variáveis, os componentes principais. Estes componentes principais foram utilizados como variáveis de entrada da regressão múltipla por componentes principais e das redes neuronais artificiais. Na regressão por mínimos quadrados parciais os dados originais foram usados como variáveis de entrada. Tomando em consideração o desempenho dos modelos desenvolvidos, com recurso a parâmetros estatísticos, foi possível concluir que a regressão múltipla por componentes principais contribuiu para piores desempenhos. Melhores desempenhos foram obtidos com a aplicação da regressão por mínimos quadrados parciais e das redes neuronais artificiais. No entanto, foi com os modelos de redes neuronais artificiais que melhores desempenhos foram obtidos em quase todas as propriedades analisadas. Tendo em conta os resultados obtidos, foi possível concluir que a espectroscopia de ressonância magnética nuclear combinada com métodos estatísticos multivariáveis pode ser usada para prever as propriedades físico-químicas de derivados de petróleo. Demonstrou-se que esta técnica pode ser considerada como uma potencial alternativa aos métodos convencionais tendo-se obtido resultados bastantes promissores.

This thesis studies a new class of high gain planar resonant cavity antennas based on metamaterial surfaces. High-gain planar antennas are becoming increasing popular due to their significant advantages (e.g. low profile, small weight and low cost). Metamaterial surfaces have emerged over the last few years as artificial structures that provide properties and functionalities not readily available from existing materials. This project addresses novel applications of innovative metamaterial surfaces on the design of high-gain planar antennas. A ray analysis is initially employed in order to describe the beamfonning action of planar resonant cavity antennas. The phase equations of resonance predict the possibility of low-profile/subwavelength resonant cavity antennas and tilted beams. The reduction of the resonant cavity profile can be obtained by virtue of novel metamaterial ground planes. Furthermore, the EBG property of metamaterial ground planes would suppress the surface waves and obtain lower backlobes. By suppressing the TEM mode in a resonant cavity, a novel aperture-type EBG Partially Reflective Surface (PRS) is utilized to get low sidelobes in both planes (E-plane and H-plane) in a relatively finite structure. The periodicity optimization of PRS to obtain a higher maximum directivity is also investigated. Also it is shown that antennas with unique tilted beams are achieved without complex feeding mechanism. Rectangular patch antennas and dipole antennas are employed as excitations of resonant cavity antennas throughout the project. Three commercial electromagnetic simulation packages (Flomerics Microstripes ™ ver6.S, Ansoft HFSSTM ver9.2 and Designer ™ ver2.0) are utilized during the rigorous numerical computation. Related measurements are presented to validate the analysis and simulations.

abstract: Image super-resolution (SR) is a low-level image processing task, which has manyapplications such as medical imaging, satellite image processing, and video enhancement, etc. Given a low resolution image, it aims to reconstruct a high resolution image. The problem is ill-posed since there can be more than one high resolution image corresponding to the same low-resolution image. To address this problem, a number of machine learning-based approaches have been proposed. In this dissertation, I present my works on single image super-resolution (SISR) and accelerated magnetic resonance imaging (MRI) (a.k.a. super-resolution on MR images), followed by the investigation on transfer learning for accelerated MRI reconstruction. For the SISR, a dictionary-based approach and two reconstruction based approaches are presented. To be precise, a convex dictionary learning (CDL) algorithm is proposed by constraining the dictionary atoms to be formed by nonnegative linear combination of the training data, which is a natural, desired property. Also, two reconstruction-based single methods are presented, which make use of (i)the joint regularization, where a group-residual-based regularization (GRR) and a ridge-regression-based regularization (3R) are combined; (ii)the collaborative representation and non-local self-similarity. After that, two deep learning approaches are proposed, aiming at reconstructing high-quality images from accelerated MRI acquisition. Residual Dense Block (RDB) and feedback connection are introduced in the proposed models. In the last chapter, the feasibility of transfer learning for accelerated MRI reconstruction is discussed.
Dissertation/Thesis
Doctoral Dissertation Computer Science 2020

碩士
國立臺灣大學
工程科學及海洋工程學研究所
103
Noise is one of the main sources of quality deterioration not only for visual inspection but also in computerized processing in brain magnetic resonance (MR) image analysis such as tissue classification, segmentation and registration. Accordingly, noise removal in brain MR images is important for a wide variety of subsequent processing applications. However, most existing denoising algorithms require laborious tuning of parameters that are often sensitive to specific image features and textures. Automation of these parameters through artificial intelligence techniques will be highly beneficial. In the present study, an artificial neural network associated with texture feature analysis is proposed to establish a predictable parameter model and automate the denoising procedure. In our approach, a large number of image attributes are extracted based on four categories: 1) Basic image statistics. 2) Gray-level co-occurrence matrix (GLCM). 3) Gray-level run-length matrix (GLRLM) and 4) Tamura texture features. Based on the t-test and the sequential forward floating selection (SFFS) methods, the optimal texture features are selected and incorporated into a back propagation neural network system. We have used a wide variety of simulated T1-weighted MR images and clinical images to evaluate the proposed automatic denoising system. Experimental results indicated that the proposed method accurately predicted the bilateral filtering parameters and automatically removed the noise in a number of MR images with satisfactory quantity and quality.

Alzheimer’s disease (AD) is a neurodegenerative illness and is considered one of the main causes of dementia affecting millions of people. In its early stage – Mild-cognitive impairment (MCI) – is asymptomatic. Furthermore, although several studies have been made, until this day, no cure is yet available. Currently, some pharmaceuticals provide a slowing of symptoms if administered in early stages. However, to do so, the diagnosis needs to be properly performed to distinguish AD different stages. Thereby, there remains a growing need for early diagnosis to minimise AD impact by delaying it and its underlying effects. This work main purpose is to create an intelligent system that enables Alzheimer’s automatic detection using Magnetic Resonance Imaging (MRI). To do so, a set of MRI images were analysed in the sagittal, coronal, and axial anatomical views and certain features were extracted and pre-selected to feed machine learning classic algorithms and a deep learning algorithm. On the one hand, for the Machine Learning classic algorithms, and for the comparison between: (1) AD vs Control (CN), a Bagged Trees Classifier reached a discrimination accuracy of 93.3!; (2) AD vs MCI, Quadratic SVM classifier got a discrimination accuracy of 87.7!; (3) CN vs MCI, Fine KNN and Subspace KNN classifiers achieved a discrimination accuracy of 88.2!, respectively; and (4) All vs All, the Subspace KNN classifier provided a discrimination accuracy of 75.3!. On the other hand, for the Deep Learning algorithm, and for the comparison between: (1) AD vs CN, a discrimination accuracy of 82.2! was achieved; (2) AD vs MCI, got a discrimination accuracy of 75.4!; (3) CN vs MCI, reached a discrimination accuracy of 83.8!; and (4) All vs All, reached a discrimination accuracy of 64.0!. In the CN vs MCI comparison, the proposed method, when compared with methods that use structural MRI (sMRI), showed an increase in classification accuracy of 9%. Therefore, the potential of this work in the diagnosis of AD, mainly in its early stages, is reinforced.
A doença de Alzheimer (DA) é uma doença neurodegenerativa que afeta milhões de pessoas, sendo considerada uma das principais causas de demência. A sua fase inicial - défice cognitivo ligeiro (DCL) – caracteriza-se por ser assintomática, e embora vários estudos tenham sido realizados, atualmente, ainda não existe uma cura disponível. No entanto, existem alguns medicamentos que proporcionam redução dos sintomas se administrados nas fases iniciais da doença. Contudo, para isto ser possível, o diagnóstico necessita de ser realizado corretamente, distinguindo-se as diferentes fases da doença. Deste modo, subsiste uma necessidade crescente de efetuar um diagnóstico precoce para minimizar o impacto da doença de Alzheimer, atrasando-a, bem como aos efeitos que lhe são subjacentes. Este trabalho tem como principal objetivo conceber um sistema inteligente que permita a deteção automática da doença de Alzheimer utilizando imagens de ressonância magnética (RM). Para tal, analisou-se um conjunto de imagens nos planos anatómicos sagital, frontal e horizontal, tendo sido extraídas e pré-selecionadas determinadas características para alimentar diferentes algoritmos clássicos de Machine Learning e de Deep Learning. Por um lado, para a técnica de Machine Learning clássica, e para a comparação entre: (1) Controlo (CN) vs AD, o classificador Bagged Trees atingiu uma precisão de discriminação de 93,3!; (2) MCI vs AD, o classificador Quadratic SVM obteve uma precisão de discriminação de 87,7!; (3) CN vs MCI, os classificadores Fine KNN e Subspace KNN atingiram uma precisão de discriminação de 88,2!, respetivamente; e (4) Todos vs Todos, o classificador Subspace KNN proporcionou uma precisão de discriminação de 75,3!. Por outro lado, para a técnica de Deep Learning, e para a comparação entre: (1) CN vs AD, foi alcançada uma precisão de discriminação de 82,2!; (2) MCI vs AD, obteve uma precisão de discriminação de 75,4!; (3) CN vs MCI, atingiu uma precisão de discriminação de 83,8!; e (4) Todos vs Todos, alcançou uma precisão de discriminação de 64,0!. Na comparação CN vs MCI, o método proposto, quando comparado com métodos que recorrem a RM estrutural, mostrou um aumento na precisão de classificação de 9%. Por conseguinte, é reforçado o potencial deste trabalho no diagnóstico da DA, principalmente nas suas fases iniciais.

This thesis studies various effects based on the excitation of surfaces plasmons in various plasmonic nanostructures. We start the thesis with a general introduction of the field of plasmonics in Chapter 1. In this chapter we discuss both propagating surface plasmon polaritons (SPPs) and localized surface plasmons (LSPs), how they are related to each other through the Bohr condition, the features of subwavelength confinement and near-field enhancement, and wave guidance through coupled LSPs. Then after the discussion of the achievements and challenges in this field (Section 1.3) we will outline the basic structure of the thesis at the end of this chapter (Section 1.4). In Chapter 2 we demonstrate a new mechanism to achieve complete spectral gap without periodicity along propagation direction based on the coupling of backward and forward modes supported by plasmonic nanostructures. We study the backward modes in single cylindrical plasmonic structures (Section 2.2) and focus on the two simplest cases: nanowires and nanocavities. Afterwards, we demonstrate how to achieve spectral gaps in coupled plasmonic nanocavities (Section 2.3). A polarization-dependent spectral gap is achieved firstly in two coupled nanocavities which support forward and backward modes respectively (Section 2.3.1). At the end we demonstrate a complete spectral gap, which is induced by the symmetry of a four-coupled-nanocavity system (Section 2.3.2). In Chapter 3 we study beam shaping in plasmonic potentials. Based on the similarity between Schrodinger equation for matter waves and paraxial wave equation for photons, we introduce the concept of plasmonic potentials and demonstrate how to obtain different kinds of potentials for SPPs in various modulated metal-dielectric-metal (MDM) structures. We investigate firstly the parabolic potentials in quadratically modulated MDM and the beam manipulations in such potentials, including polychromatic nanofocusing in full parabolic potentials (Section 3.2.1), plasmonic analogue of quantum paddle balls in half parabolic potentials (Section 3.2.2), and adiabatic nanofocusing in tapered parabolic potentials (Section 3.2.3). In the following section (Section 3.3) we show the existence of linear plasmonic potentials in wedged MDM and efficient steering of the Airy beams in such potentials (Section 3.3.2) after a brief introduction on Airy beams in free space (Section 3.3.1). In Chapter 4 we study scattering engineering by magneto-dielectric core-shell nanostructures. The introduction part (Section 4.1) gives a brief overview on the scattering of solely electric dipole (ED) or magnetic dipole (MD), and how the coexistence and interference of the ED and the MD can bring extra flexibility for scattering shaping. Afterwards, we discuss the scattering shaping by core-shell nanostructures through the interferences of electric and artificial magnetic dipoles (Section 4.2), including two examples of broadband unidirectional scattering by core-shell nanospheres (Section 4.2.1) and efficient scattering pattern shaping of core-shell nanowires (Section 4.2.2). At the end of this chapter we demonstrate polarization independent Fano resonances in arrays of core-shell nanospheres (Section 4.3.2). At the end of this thesis (Chapter 5) we summarize the results and draw the conclusions. We also discuss the challenges and possible future developments of the field of plasmonics.

The interactions of the radio frequency (RF) fields in magnetic resonance imaging (MRI) with a human body model are investigated. In particular, the interactions of these fields with an implanted pacemaker are studied. The specific absorption rate (SAR) levels in a heterogeneous body model are evaluated in two different birdcage coils - a resonant and a non-resonant coil - at a magnetic field of 1.5 T. The enhancement of the SAR due to an implanted cardiac pacemaker and the effect of the conductivity of the pacemaker lead on the SAR levels are investigated. The finite difference time domain (FDTD) technique is used to model these interactions. The SAR levels are found to be low in the heart region, and thus the SAR enhancement due to the pacemaker lead is relatively low. Modeling of the pacemaker leads as perfect conductors results in greater SAR enhancements than those produced by actual conductive leads.

The goal of this project was the development of a computer-based system using artificial intelligence to classify the limonoids, protolimonoids and triterpenoids isolated from the family Meliaceae by the Natural Products Research Group of the University of Natal, Durban. A database of samples was obtained between 1991 and 1996, part of which time the author was a member of the group and isolated compounds from Turraea obtusifolia and Turraea floribunda. Over and above the problem of complexity and similarity in structures of the above mentioned natural products, are other difficulties. These include very small amounts of sample being isolated producing very weak peak signals in the C-13 NMR spectra, extraneous peaks in the NMR spectra due to different impurities and instrument noise, non-reproducible spectra due to the pulsed Fourier transform intervals and the nuclear Overhauser effect, impure samples often isolated as stereoisomeric mixtures or as mixed esters and superposition of peak signals in the NMR spectra due to carbons in the same environment within the same compound. These factors make identification by traditional computational and expert systems impossible. As a result of these shortcomings, the author has developed a novel approach using artificial neural network techniques. The artificial neural network system developed used real data from the 300 MHz NMR spectrometer in the Department of Chemistry, Durban. The system was trained to discriminate between limonoids, triterpenoids and flavonoids/coumarins from the C-13 NMR spectra of pure, impure and unseen compounds with an accuracy of better than 90%. Further differentiation of the glabretals from the rest of the protolimonoids as well as from the rest of the triterpenoids showed similarly significant results. Finally, individual limonoid discrimination within the limonoid dataset was extremely successful. Apart from its application to the extractives from Meliaceae, the methodology and techniques developed by the author can be applied to other sets of extractives to provide a robust method for the spectral classification of pre-identified natural products.
Thesis (Ph.D.)-University of Natal, Durban, 1998.

Deep learning has overwhelmingly impacted post-acquisition image-processing tasks, however, there is increasing interest in more tightly coupled computational imaging approaches, where models, computation, and physical sensing are intertwined. This dissertation focuses on how to leverage the expressive power of deep learning in image reconstruction. We use deep learning in both the sensor data domain and the image domain to develop new fast and efficient algorithms to achieve superior quality imagery. Metal artifacts are ubiquitous in both security and medical applications. They can greatly limit subsequent object delineation and information extraction from the images, restricting their diagnostic value. This problem is particularly acute in the security domain, where there is great heterogeneity in the objects that can appear in a scene, highly accurate decisions must be made quickly, and the processing time is highly constrained. Motivated primarily by security applications, we present a new deep-learning-based MAR approach that tackles the problem in the sensor data domain. We treat the observed data corresponding to dense, metal objects as missing data and train an adversarial deep network to complete the missing data directly in the projection domain. The subsequent complete projection data is then used with an efficient conventional image reconstruction algorithm to reconstruct an image intended to be free of artifacts. Conventional image reconstruction algorithms assume that high-quality data is present on a dense and regular grid. Using conventional methods when these requirements are not met produces images filled with artifacts that are difficult to interpret. In this context, we develop data-domain deep learning methods that attempt to enhance the observed data to better meet the assumptions underlying the fast conventional analytical reconstruction methods. By focusing learning in the data domain in this way and coupling the result with existing conventional reconstruction methods, high-quality imaging can be achieved in a fast and efficient manner. We demonstrate results on four different problems: i) low-dose CT, ii) sparse-view CT, iii) limited-angle CT, and iv) accelerated MRI. Image domain prior models have been shown to improve the quality of reconstructed images, especially when data are limited. A novel principled approach is presented allowing the unified integration of both data and image domain priors for improved image reconstruction. The consensus equilibrium framework is extended to integrate physical sensor models, data models, and image models. In order to achieve this integration, the conventional image variables used in consensus equilibrium are augmented with variables representing data domain quantities. The overall result produces combined estimates of both the data and the reconstructed image that is consistent with the physical models and prior models being utilized. The prior models used in both image and data domains in this work are created using deep neural networks. The superior quality allowed by incorporating both data and image domain prior models is demonstrated for two applications: limited-angle CT and accelerated MRI. A major question that arises in the use of neural networks and in particular deep networks is their stability. That is, if the examples seen in the application environment differ from the training environment will the performance be robust. We perform an empirical stability analysis of data and image domain deep learning methods developed for limited-angle CT reconstruction. We consider three types of perturbations to test stability: adversarially optimized, random, and structural perturbations. Our empirical analysis reveals that the data-domain learning approach proposed in this dissertation is less susceptible to perturbations as compared to the image-domain post-processing approach. This is a very encouraging result and strongly supports the main argument of this dissertation that there is value in using data-domain learning and it should be a part of our computational imaging toolkit.

Diffusion Tensor Imaging (DTI) is a magnetic resonance imaging (MRI)-based technique that has frequently been used for the identification of brain biomarkers of neurodevelopmental and neurodegenerative disorders because of its ability to assess the structural organization of brain tissue. In this work, I present (1) preclinical findings of a longitudinal DTI study that investigated asymptomatic high school football athletes who experienced repetitive head impact and (2) an automated pipeline for assessing the quality of DTI images that uses a convolutional neural network (CNN) and transfer learning. The first section addresses the effects of repetitive subconcussive head trauma on the white matter of adolescent brains. Significant concerns exist regarding sub-concussive injury in football since many studies have reported that repetitive blows to the head may change the microstructure of white matter. This is more problematic in youth-aged athletes whose white matter is still developing. Using DTI and head impact monitoring sensors, regions of significantly altered white matter were identified and within-season effects of impact exposure were characterized by identifying the volume of regions showing significant changes for each individual. The second section presents a novel pipeline for DTI quality control (QC). The complex nature and long acquisition time associated with DTI make it susceptible to artifacts that often result in inferior diagnostic image quality. We propose an automated QC algorithm based on a deep convolutional neural network (DCNN). Adaptation of transfer learning makes it possible to train a DCNN with a relatively small dataset in a short time. The QA algorithm detects not only motion- or gradient-related artifacts, but also various erroneous acquisitions, including images with regional signal loss or those that have been incorrectly imaged or reconstructed.

The first step of photosynthetic metabolism effects the facile oxidation of water to dioxygen and hydrogen cations. This is achieved through an incompletely-understood process of light-driven four-electron oxidation at the Mn4CaO5 cofactor of the Oxygen Evolving Complex (OEC) of the Photosystem II (PSII) holoenzymatic complex in photosynthetic autotrophs. Biomimesis of this reaction—artificial photosynthesis—may offer energy-efficient routes to industrial hydrogen generation and value-added derivatives, with implications for solar energy fixation. This thesis consists of a compilation of four publications relating to Density Functional Theory (DFT) studies of structural and spectroscopic aspects of the OEC of PSII. These publications consist of research resolving the basis of structural anomalies in metal-substituted PSII, combinatoric simulation of difference spectra corresponding to proton-coupled oxido-reduction scenarios of PSII models, simulation of the hyperfine and superexchange magnetic interactions in PSII models, and the development of a methodology for obtaining vibrational intensities in the Mobiel Block Hessian (MBH) approximation, with applications to accelerated modeling of the vibrational structure of complex models of PSII, as well as other large molecules. These publications are presented alongside explanatory introductions and preceded by a general survey of the state of the art of photosynthesis research, context for the relevance of this research, and methodological discussion. Concluding remarks are also presented.