Indiana University-Purdue University Indianapolis (IUPUI) / This thesis presents a Field Programmable Gate Array (FPGA) based embedded system which is used to achieve high speed segmentation of 3D images. Segmenta- tion is performed using Expectation-Maximization with Maximization of Posterior Marginals (EM/MPM) Bayesian algorithm. In this system, the embedded processor controls a custom circuit which performs the MPM and portions of the EM algorithm. The embedded processor completes the EM algorithm and also controls image data transmission between host computer and on-board memory. The whole system has been implemented on Xilinx Virtex 6 FPGA and achieved over 100 times improvement compared to standard desktop computing hardware.

We present a study that investigates user performance benefits of 3D stereo in modern video games. Based on an analysis of several video games that are best suited for use with commercial 3D stereo drivers and vision systems, we chose five modern titles focusing on racing, first person shooter, third person shooter, and sports game genres. For each game, quantitative and qualitative measures were taken to determine if users performed better and learned faster in the experimental group (3D stereo display) than in the control group (2D display). A game experience pre-questionnaire was used to classify participants into beginner, intermediate, and advanced gameplay categories to ensure prior game experience did not bias the experiment. Our results indicate that even though participants preferred playing in 3D stereo, for the games we tested, it does not provide any significant advantage in overall user performance. In addition, users‟ learning rates were comparable in the 3D stereo display and 2D display cases

Radiotherapy is a prominent cancer treatment modality in medicine, aiming to deliver adequate doses to the target while minimizing harm to healthy tissue. Recent advancements in computer technology, machine engineering, and imaging have facilitated intricate treatment planning and accurate radiation administration. These advancements have allowed for more precise dose distributions to be delivered to cancer patients. However, even small discrepancies in setup or delivery can result in significant dose variations. While treatment planning systems provide 3D dose calculations, there is currently a lack of 3D measurement tools in the clinic to verify the accuracy of dose calculation and delivery. Presently, medical physicists rely on 2D dose plane comparisons with treatment planning calculations using gamma index analyses. However, these results do not directly correlate with clinical dose-volume constraints, and detecting delivery errors using 1D or 2D dosimetry is challenging. The implementation of 3D dosimetry not only ensures the safety of radiation treatment but also facilitates the development of new emerging radiation treatment techniques. This study aims to commission and validate a clinically viable optical scanner for 3D dosimetry and apply the developed system to address current clinical and pre-clinical challenges, thereby advancing our understanding of treatment uncertainties in modern radiotherapy. The optical CT scanner that was developed comprises four key components: an LED illuminator, an aquarium with matching fluid, a fiber optic taper, and a CCD camera. The LED illuminator emits uniform and parallel red light at a peak wavelength of 625 nm and a full width at half maximum (FWHM) of 20 nm in continuous mode. The aquarium is constructed with transparent acrylic walls and is designed to accommodate the 3D dosimeter PRESAGE, which can be fixed on a rotation stage inside the tank. Clear acrylic has excellent optical clarity and light transmission, with a refractive index of 1.49 that is close to the average refractive index (1.54) of PRESAGE. To match the refractive index of the 3D dosimeters, a matching liquid composed of 90% Octyl Salicylate and 10% Octyl-P-Methoxy Cinnamate is filled in the tank. The fiber optic taper serves two functions: first, it demagnifies the projection images while preserving their shape, and second, it effectively reduces the acceptance angle of the light reaching the CCD camera. The CCD camera used in the system is an Allied Vision model with a resolution of 0.016 mm, capable of acquiring 2D projection images from various angles. The principle of the optical CT scanner follows that of CT imaging, where 2D projection images from different angles are used to reconstruct volumetric 3D dose images using the filtered back projection technique. To validate the dosimetric measurements and assess the uncertainties of the 3D dosimetry system, 21 benchmark experiments, including mechanical, imaging, and dosimetry tests were conducted. Furthermore, the developed system was employed for various applications, including patient-specific IMRT QA, small field dosimetry using kilovoltage and megavoltage beams, as well as end-to-end testing of stereotactic radiosurgery. A comprehensive analysis assessed uncertainties in each scanner component. Mechanical tests showed maximum uncertainties below 1%. By employing background subtraction and calibration techniques, measurement uncertainty was reduced to <1% in the optimal dose range. Background subtraction resulted in a remarkable 77% reduction in uncertainty by mitigating artifacts, ambient light, and refractive light. Reproducibility was excellent, with mean and standard deviation of dose differences below 0.4% and 1.1%, respectively, in three repeat scans. Dose distribution measurements exhibited strong agreement (passing rates: 98%-100%) between 3D measurements, treatment planning calculations, and EBT3 film dosimetry. Results confirm the optical CT scanner's robustness and accuracy for clinical 3D radiation dosimetry. The study also demonstrates that the developed 3D dosimetry system surpasses the limitations of traditional 2D gamma tests by providing clinicians with more clinically relevant information. This includes measured dose-volume histograms (DVHs) and the evaluation of gamma failing points in 3D space, enabling a comprehensive assessment of individual treatment plans. Furthermore, the study showcased the feasibility of utilizing this system to characterize a radiosurgery platform. It successfully assessed mechanical and dosimetric errors in off-axis delivery and evaluated the accuracy of treatment planning dose calculations, including modeling small fields, out-of-field dose, and multi-leaf collimator (MLC) characteristics. In addition, compelling evidence was presented that the high-resolution 3D dosimeter used in this study is capable of accurate dosimetry for both megavoltage and kilovoltage small fields. Importantly, the dosimeter exhibits no energy or dose rate dependence, further supporting its reliability and suitability for precise dosimetry measurements. The intricate and three-dimensional nature of dose distributions in modern radiotherapy necessitated the development of 3D dosimetry measurements, particularly for treatments with precise margins, such as SRS and SBRT. The newly developed 3D dosimetry system offers significant enhancements to current QA practices, delivering more clinically relevant comparison results and bolstering patient safety. Furthermore, it can be utilized for independent inspections across multiple institutions or remote dosimetry verification. Beyond its applications in clinical settings, the presented 3D dosimetry system holds the potential to expedite the development and utilization of novel radiation platforms.

How does the brain integrate information from individual neurons? One efficient way to investigate systematic neuroscience is to record the whole brain down to singular neuron level. Caenorhabditis elegans, a 1 mm long, transparent nematode species, is ideally suited as a starting point. Every C. elegans hermaphrodite has a fixed set of 302 neurons. All neuron connections have been fully characterized by electron microscopy. Despite its small and simple nervous system, C. elegans exhibits a wide range of behaviors ranging from foraging, sleep to sexual activity. Recently, Yemini et al. genetically engineered a C. elegans strain where each neuron can be uniquely identified by its color code. This greatly facilitates comparison of neural recordings with literature as well as underlying connectomics. However, it is a daunting task to record the whole nervous system at cellular resolution of a freely moving worm. The imaging system needs to achieve high 3D imaging speed (10+ volumes per second) to avoid motion blur while also maintaining single cell resolution and reasonable field of view.Over the past decade, light sheet microscopy has emerged as a promising technique with great spatial resolution and reduced phototoxicity. Swept, confocally-aligned, planar excitation (SCAPE) microscopy, a single objective light sheet modality developed by Hillman lab, has the advantage of an open top geometry and fast 3D imaging speed. In this proposal, I detail my work towards imaging and tracking the whole C. elegans nervous system at cellular resolution using SCAPE and the NeuroPAL strain. The first chapter introduces fundamental concepts that link the microscopy field with the C. elegans community. The second chapter involves building a new SCAPE system that incorporates new optical components and a high-speed intensified camera. The goal is to construct a workhorse system capable of capturing real-time volumetric recordings with improved resolution. The improvements stem from an improved optical design as well as careful selection of magnification and scan parameters While the new imaging system is capable of capturing high-speed volumetric images of freely moving NeuroPAL worms with single-cell resolution, there is no suitable neuron tracking algorithm to robustly extract neural activities from the data. Indeed, the density of the neurons as well as the vigorous movement of the worm is unprecedented. Chapter 3 and 4 constitute two parts of a broader neuron tracking algorithm. In Chapter 3, I introduce an iterative neural network based algorithm for unsupervised 3D image registration. In Chapter 4, a Gaussian Mixture Model based algorithm is proposed that simulates the raw data as the mixture of 3D Gaussian functions. Chapter 5 is the finale where I integrate of all proposed imaging and tracking methods in recording neural activity from the whole nervous system in freely-behaving NeuroPAL worms. Three applications are demonstrated, which spans from whole nervous system recording to investigation of class-dependent ventral nerve cord motor neurons during locomotion. In Chapter 6, I report progress towards building the next-generation SCAPE with higher resolution/collection efficiency. A custom-designed zero working distance objective is demonstrated, which uses off-the-shelf objective with novel refractive-index-matched material to achieve high collection numerical aperture without sacrificing field of view (FOV).

The primary purpose of the thesis is to discuss the use of Magnetic Resonance Imaging (MRI) in functional proton radiosurgery. The methods presented in this thesis were specifically designed to correct gradient nonlinearity distortion, the single greatest hurdle that limits the deployment of MRI-based functional proton radiosurgery systems. The new system central in the thesis fully utilized MRI to provide localization of anatomical targets with submillimeter accuracy. The thesis provides analysis and solutions to the problems related to gradient nonlinearity distortion. The characteristics of proton radiosurgery are introduced, in addition to a discussion of its advantages over other current methods of radiation oncology. A historical background for proton radiosurgery is also presented, along with a description of its implementation at Loma Linda University Medical Center (LLUMC), where a new system for functional proton radiosurgery has been proposed and is currently under development.

Traditionally, surface topography measurement was in the domain of quality control of engineering parts. With the advancement of manufacturing technology and affordable computational costs, different types of surfaces are produced with varied shapes and surface textures. These pose significant measurement problems, therefore, surface topography research is gaining momentum to achieve a better control of the surface. Coherence scanning interferometry (CSI) is one of the most common techniques used for measurement of surface topography. It is preferred over tactile and other non-contact techniques since it provides fast and accurate measurement with high vertical (~ 1 nm) and lateral (~1 μm) resolutions over larger areas without any damage to the surface. Essentially, CSI is treated as one dimensional (1D) superposition of the light waves from an object and a reference that generates a three dimensional (3D) interferogram. Secondly, despite the advantages, there is no standard configuration of CSI that can provide absolute surface topography measurement of an engineering part with multiple materials. An effective solution to this problem will be particularly useful in the field of semiconductor and bio-related industries where chips and instruments are made of many materials. In this Thesis, first, the CSI technique is analysed in terms of a wider theoretical framework of 3D linear filtering technique which shows the similarities among other seemingly disparate techniques such as confocal and optical coherence tomography. Due consideration to the spectral characteristic of the source and the effect of numerical aperture are given and important parameters such as vertical and lateral resolutions are computed to compare this theory with standard analysis methods. Additionally, it is shown that the 3D fringe pattern can be considered to be a superposition of a reference field and the scattered field from the top foil-like layer on the top the object. The scattered field from this foil object is dependent on the normal Fresnel reflection coefficients. Therefore, it explains the phase offset and the proportional height offset introduced by different materials, especially, metals. In an object, where multiple materials are present, each material introduces different phase to the fringe pattern and therefore, the surface topography of the entire object is altered. To overcome this problem, the optical polarising properties of the material are exploited. A novel configuration of polarisation sensitive CSI is presented where interferograms with orthogonal circular polarisations are recorded and analysed. The configuration, initially, needs to be calibrated with a material and after that at each point on the object, the refractive index and height offset can be calculated. Therefore, it can be dually used to identify unknown materials present on the object and also to compensate for the height offset introduced by each material to produce absolute surface topography of the entire object. The configuration provides good agreement with ellipsometric results for metals. Additionally, it retains the advantages of high vertical and lateral resolution same as other standard coherence scanning interferometers.

L’objet de cette étude est une analyse de 32 individus néolithiques représentés par des crânes et/ou des mandibules mis au jour au Mali et en Mauritanie. Les individus composant l’échantillon ostéoarchéologique sont issus de cinq sites présentant des différences géographiques et chronologiques. En dépit de ces différences, nous nous sommes interrogés sur les liens qui pouvaient exister entre eux. Cette analyse, qui a consisté à reconsidérer les caractères morphologiques des crânes et/ou mandibules, leur variabilité, s’est appuyée sur une technologie nouvelle : l’imagerie médicale et tridimensionnelle. Notre premier objectif a été de tenter de répondre à la question suivante : dans quelle mesure l’anthropologie biologique peut-elle s’enrichir des nouvelles technologies issues du domaine de l’imagerie médicale ? Notre questionnement ne s’est pas arrêté pas à cette perspective méthodologique. En effet, l’échantillon ostéoarchéologique présente une particularité morphologique frappante : l’aspect très robuste présenté par de nombreux individus. Cette robustesse confère un caractère archaïque aux sujets par bon nombre de caractères, et ceci malgré des datations néolithiques. Le second objectif de ce travail a été d’ordre paléoanthropologique et a reposé sur l’étude de cette robustesse. La méthode mise en place a permis d’obtenir 146 mesures par individu (en fonction de l’état de fragmentation de chacun). Il s’est avéré que, non seulement, ce caractère a été constaté, mais qu’il semble plus précisément lié à un élément anatomique, la mandibule, ce qui pourrait être le résultat d’une adaptation de l’appareil manducateur aux contraintes environnementales et alimentaires. / The object of this study is an analysis of 32 neolithic individuals represented by skulls and/or mandibles brought to light in Mali and in Mauritania. The individuals composing the osteoarcheological sample arise from five sites presenting geographical differences and chronological. In spite of these differences, we wondered about the links which could exist between them. This analysis, which consisted in reconsidering the morphological characters of skulls and/or mandibles, their variability, leaned on a new technology : the medical and three-dimensional imaging. Our first objective was to try to answer the following question : to what extent the biological anthropology can increase from new technologies stemming from the domain of the medical imaging ? Our questioning did not stop at in this methodological perspective. Indeed, the osteoarcheological sample presents a striking morphological peculiarity: the very strong aspect presented by numerous individuals. This robustness confers an archaic character on the subjects by a lot of characters, and this in spite of neolithic datings. The second objective of this work was of order paleoanthropological and based on the study of this robustness. The organized method allowed to obtain 146 measures by individual (according to the state of fragmentation of each). It appears that, not only this character was noticed, but it seems more exactly being bound to an anatomical element, the mandible, what could be the result of an adaptation of the manducative system to the environmental and food constraints.

O cirurgião-dentista necessita muitas vezes visualizar imagens associadas a estruturas importantes onde um bom conhecimento da anatomia crânio facial e dentomaxilar é imprescindível na elaboração de um diagnóstico correto. O seio maxilar é uma estrutura anatômica localizada no terço médio da face, sendo de extrema importância por estar localizado muito próximo às estruturas dentárias posteriores, devendo fazer parte do conhecimento dos cirurgiões-dentistas, nas diversas especialidades. Nesse estudo foram avaliados 50 exames de Tomografia Computadorizada de Feixe Cônico (TCFC) onde os seios maxilares podiam ser visualizados, resultando em um total de 78 extensões para diferentes regiões foram encontradas, 17 indivíduos apresentavam espessamento da membrana sinusal, 14 mostravam imagens compatíveis com cistos de retenção de muco e 6 apresentavam velamento do seio maxilar, variando entre total e parcial. Um total de 64 septos foram identificados no interior da cavidade sinusal, 86 cúpulas alveolares localizadas nas raízes dos diferentes dentes, 126 raízes foram encontradas em íntimo contato com o assoalho sinusal. Os resultados foram estatisticamente significantes para a relação entre cúpulas alveolares e extensões alveolares para a raiz palatina do dente 26, a presença de septos no teto dos seios maxilares de mulheres e da relação da média da região desdentada do dente 26 para os homens, com os valores de p < 0,05. Diversos fatores relacionados à morfologia do osso maxilar, a ausência de dentes e a própria conformação da cavidade sinusal podem resultar em diferentes formatos entre indivíduos e variações até no mesmo indivíduo. O exame por Tomografia Computadorizada de Feixe Cônico (TCFC) dos seios maxilares permite uma avaliação tridimensional e sua relação com as estruturas adjacentes, ou seja, uma perfeita reprodução da região dentomaxilofacial, auxiliando de maneira precisa os profissionais que irão intervir junto ou próximo a essa estrutura anatômica. / Dentistry professionals are required to analyze, in many cases, images associated to important structures where a great knowledge of the dentomaxillofacial anatomy is vital for an accurate diagnostic. The maxillary sinus is an anatomical structure that occupies the upper 2/3s of the maxillary bone and it is very important because the proximity with the posterior dental structures, it must be well known by dentists in the several specialties. In this study, 50 Cone Beam Computed Tomography (CBCT) exams in which the maxillary sinus could be evaluated, showed a total of 78 extension toward different regions were achieved, 17 subjects displayed a mucosal thickening, 14 evidenced a compatible imaging of mucous retention cysts and 6 presented a partially or completely filled maxillary sinus. A total of 64 maxillary sinus septa were identified, 86 different roots were localized inside de sinus, 126 roots were found to be in proximity with the sinus floor. The results were statistically significant to the relation made between these roots, which were inside the sinus and the alveolar extensions to the palatine root of the tooth 26, the presence of maxillary sinus septa located in the roof of the women maxillary sinus and the relation made between the median value for men in the edentulous region of the tooth 26, with a p value < 0,05. Its anatomy is one of the most varied, and many factors related to the maxillary bone morphology like edentulousness and the sinus conformation can result on different shapes, even in the same person. The Cone Beam Computed Tomography (CBCT) of the maxillary sinus provides a three-dimensional evaluation and its relationship with the boundary structures. It is a precise reproduction of the dentomaxillofacial region, providing valuable support to the professional who will work on or next to this anatomical structure.

Horizontal scanning instruments, such as, atomic force microscopes and scanning laser microscopes, acquire three-dimensional topographic maps of surfaces, at scales ranging from tenths of nanometers to hundreds of millimeters, by measuring elevations along a series of traces scanning a region of the surface. Random and systematic errors may influence parameters calculated from these topographic maps. This work investigates anisotropic artifacts in atomic force microscope and a scanning laser microscope measurements by looking at difference between parameters calculated in the tracing and scanning directions. It is found that horizontal scanning profiling instruments systematically introduce anisotropic measurement artifacts when measuring both isotropic and anisotropic surfaces.

A level set-based segmentation procedure has been implemented to identify target object boundaries from 3D medical ultrasound images. Several test images (simulated, scanned phantoms, clinical) were subjected to various preprocessing methods and segmented. Two metrics of segmentation accuracy were used to compare the segmentation results to ground truth models and determine which preprocessing methods resulted in the best segmentations. It was found that by using an anisotropic diffusion filtering method to reduce speckle type noise with a 3D active contour segmentation routine using the level set method resulted in semi-automated segmentation on par with medical doctors hand-outlining the same images.