The spread of the pulmonary disease among humans is a very rapid process and it stands as the third highest killer in the United States of America. Computed Tomography (CT) scanning allows us to obtain detailed images of the pulmonary anatomy including the airways. The complexity of the tree makes the process of manual segmentation tedious, time-consuming, and variant across individuals. The resultant airway segmentation, whether arrived at manually or through the aid of computers, can then be used to measure airway geometry, study airway reactivity, and guide surgical interventions. The thesis addresses these problems and suggests a fully automated technique for segmenting the airway tree in three-dimensional (3-D) micro-CT images of the thorax of an ex-vivo mouse. This novel technique is a several step approach consisting of: 1. The feature calculation of individual voxels of the micro-CT image, 2. Selection of the best features for classification (obtained from 1), 3. KNN-classification of voxels by the best selected features (from 2) and 4. Region growing segmentation of the KNN classified probability image. KNN classification algorithm has been used for the classification of the voxels of the image (into airway and non-airway voxels) based on the image features, the results of which have then been processed using the region growing segmentation algorithm to obtain the final set of results for segmentation. The segmented airway of the ex-vivo mouse lung volume can be analyzed using a commercial software package to obtain the measurements.

Among the different radiodiagnostic techniques, computed tomography (CT) represents the major contribution to the collective dose. In order to have available a tool to assess and ascertain the imparted dose upon CT examinations, we have developed and validated a method to calculate radiation absorbed dose in CT examinations from images of phantoms and standard patients by using a voxel-based Monte Carlo simulation method.Simulations and measurements of radiation dose were conducted in air, within cylindrical phantoms of both circular and elliptic cross sections, and in an anthropomorphic phantom, with the simulations being made on volumes based on their own CT images. The radiation beam geometry (focus-to-axis distance, field of view, collimation, and primary and beam shaper filtration) and the X-ray spectral distribution for each scanner under study were implemented for the simulation. Simulations were performed for both single axial scan and a series of contiguous scans covering the complete phantom-object volume. All the axial CT images were reduced to a 256 x 256 matrix and stacked in a volume. Patient images were segmented prior to simulation of radiation transport by using three categories (lung, muscle and bone). To test the method, values of CT specific dose quantities for CT exposure simulations were calculated and compared to those derived from measurements performed free-in-air and within cylindrical phantoms. To complete the scope of the work, series of CT scans of the head, neck, and trunk of an anthropomorphic phantom and standard patients were simulated. Using the EGS4 code, written in an extended FORTRAN called Mortran3, we have estimated first the radiation dose in the individual voxels; radiation transport was simulated by means of the EGS4 code system. The simulation of photon transport included incoherent and coherent scattering, photoelectric absorption with creation of fluorescent photons or Auger electrons. The photon interaction cross sections and the electronic density for each material or compound were prepared by the PEGS4 pre-processor. The electron interaction processes considered were inelastic collisions with atomic electrons, scattering effects by the nucleus, and bremsstrahlung losses that are negligible for the energies considered. The low energy cutoff for photon transport was 5 keV, whereas for electrons was fixed at 30 keV, with a ESTEPE (fraction of the energy lost in a step) value of 1%. The conversion of energy fluence to dose rate was made using the mass energy-absorption coefficients from Hubbell and Seltzer. The number of histories in each simulation was 3×106 per each 1 cm slice.Secondly, the dose CT indices for peripheral and central positions and the dose to bladder were estimated as well, the later being used to test organ doses in the anthropomorphic phantom. The dose measurements were performed by using a "pencil" ionisation chamber and 30 thermoluminiscent dosimeters. The comparison between simulated and measured dose data shows differences below 5% in all cases.The designed program allows contrasted results to be achieved, and might be an adequate tool for the analysis of doses imparted to individual patients with different models of CT scanner. / De las diferentes técnicas usadas con fines radiodiagnósticos, las exploraciones con equipos de tomografía computarizada (TC) suponen la mayor contribución a la dosis colectiva, ya que las dosis impartidas en cada exploración de TC son significativamente más elevadas que las de los estudios radiológicos convencionales. Con el fin de disponer de una herramienta para conocer las dosis impartidas a los pacientes en estudios de TC, se ha desarrollado y validado un programa de simulación para estimar la dosis absorbida en volúmenes divididos en voxels usando técnicas de Montecarlo. Para obtener los "objetos" tridimensionales divididos en voxels que representen a cada maniquí o paciente, se han transformado las propias imágenes de TC mediante la segmentación en materiales u órganos, y posteriormente se han apilado a lo largo del eje longitudinal. El haz se ha simulado teniendo en cuenta los espectros de emisión de rayos X, la filtración total del equipo (filtración primaria y filtro de forma) y reproduciendo la geometría concreta de irradiación para los tres equipos utilizados (dos helicoidales y un multicorte). Mediante el código EGS4 se han reproducido los procesos relevantes (interacción y transporte de las partículas) en el rango de energías de los equipos de TC, y se ha estimado la energía depositada en cada voxel. Durante el proceso de validación se han realizado estudios de dosis absorbida en aire, en maniquíes cilíndricos de sección circular y elíptica, y en un maniquí antropomórfico, construidos a partir de sus propias imágenes de TC. Se han comparado los resultados de la simulación con los de medidas realizadas en maniquíes físicos utilizando una cámara de ionización o dosimetría por termoluminiscencia. El análisis de las diferencias y de las incertidumbres atribuidas a los resultados ha mostrado que ambas son compatibles. Una vez validado el programa de simulación se ha aplicado a tres estudios concretos, permitiendo estimar la dosis promedio por rodaja, y la dosis promedio equivalente en distintos órganos de dos maniquíes antropomórficos y de seis pacientes seleccionados para representar las distintas morfologías de ambos sexos. También se han asignado equivalencias entre maniquíes y grupos de pacientes pediátricos de edades comprendidas entre los 0 a los 15 años. Finalmente se ha aplicado el método a un caso real de protección radiológica mediante la utilización de protectores con bismuto destinados a reducir la dosis en ojos, tiroides y mamas.El programa diseñado, que se puede adaptar a cada estudio particular, permite obtener resultados contrastados y puede ser una herramienta adecuada para analizar las dosis impartidas a pacientes individuales en cada protocolo y equipo de TC concreto. Además, el programa se muestra como herramienta eficaz para analizar los parámetros que influyen en la distribución de dosis, ya sean geométricos, del haz, o relacionados con los modos de adquisición del equipo.

The cerebral cortex is gray matter tissue which covers cerebral hemispheres. In recent years, many studies reported that abnormal cortical thickness was found in several diseases of central neural system, such as multiple sclerosis, Alzheimer's diseases, and schizophrenia. Therefore, the whole-brain measurement of cortical thickness using the non-invasive magnetic resonance imaging becomes important. However, not many algorithms were reported in the past due to the extremely complex folding structure of human cortex. In this thesis, a voxel-based cortical thickness method proposed by Hutton et al was implemented using MATLAB to achieve automated measurement. Several crucial factors, including the definition of boundary condition, interpolation method, the step size of developing each streamline, and spatial resolution of imaging space, in the implementation were discussed. In addition, the analysis of stability, or precision, of our self-developed program was evaluated . Sixteen experiments of reproducibility were performed in two months on the same 24-year-old healthy volunteer repeatedly to obtain whole-brain 3D T1WI. Cortical thickness map was calculated independently and normalized to the same coordination. Mean, standard deviation, and normalized standard deviation of 16 measurements were calculated on every cortical voxel, along with whole-brain mean cortical thickness. Various sizes of 3D smoothing kernel were applied, and the results showed stronger smoothing might help higher precision by the cost of spatial resolution.

Objective: The purpose of this study is to determine the relationship between cone beam CT (CBCT) voxel size and tooth root fracture detection. Materials and Methods: Vertical and horizontal root fractures were induced in a total of 30 teeth, and 15 teeth were left intact. Teeth were imaged with projection digital radiography and the Kodak 9000 3D CBCT system with a native voxel size of 76 μm. The CBCT voxels were then downsampled to 100 μm, 200 μm and 300 μm. Five blinded observers evaluated both sets of images with a 1 week washout interval between each set of observations. Results: CBCT outperformed the projection images for fracture detection for all voxel sizes except 300 μm (p<0.05). No significant differences were found between the different voxel sizes (p>0.05). Conclusion: Although voxel size does not impact the interpretation of root fractures, in vitro, CBCT outperformed projection imaging for voxel sizes less than 300 μm.

Objective: The purpose of this study is to determine the relationship between cone beam CT (CBCT) voxel size and tooth root fracture detection. Materials and Methods: Vertical and horizontal root fractures were induced in a total of 30 teeth, and 15 teeth were left intact. Teeth were imaged with projection digital radiography and the Kodak 9000 3D CBCT system with a native voxel size of 76 μm. The CBCT voxels were then downsampled to 100 μm, 200 μm and 300 μm. Five blinded observers evaluated both sets of images with a 1 week washout interval between each set of observations. Results: CBCT outperformed the projection images for fracture detection for all voxel sizes except 300 μm (p<0.05). No significant differences were found between the different voxel sizes (p>0.05). Conclusion: Although voxel size does not impact the interpretation of root fractures, in vitro, CBCT outperformed projection imaging for voxel sizes less than 300 μm.

The new approach of volume data visualization is presented in this paper. The proposed algorithm is based on point-based volume visualization. The main idea of it is to approximate volumetric dataset with surface made from points rather than approximate it with polygon-based iso-surface. This work covers a survey of volumetric data types, main principles of volume visualization and areas where improvements on algorithms of volume visualization are likely to be done. There is analysis of meshless iso-surface generation from multiblock data algorithm presented, suggesting dynamic iso-surface generation modification and some experiments based on these suggestions.

This thesis characterizes the problem of relative camera calibration in the context of three-dimensional volumetric reconstruction. The general effects of camera calibration errors on different parameters of the projection matrix are well understood. In addition, calibration error and Euclidean world errors for a single camera can be related via the inverse perspective projection. However, there has been little analysis of camera calibration for a large number of views and how those errors directly influence the accuracy of recovered three-dimensional models. A specific analysis of how camera calibration error is propagated to reconstruction errors using traditional voxel coloring algorithms is discussed. A review of the Voxel coloring algorithm is included and the general methods applied in the coloring algorithm are related to camera error. In addition, a specific, but common, experimental setup used to acquire real-world objects through voxel coloring is introduced. Methods for relative calibration for this specific setup are discussed as well as a method to measure calibration error. An analysis of effect of these errors on voxel coloring is presented, as well as a discussion concerning the effects of the resulting world-space error.

In recent years several depth cameras have emerged on the consumer market, creating many interesting possibilities forboth professional and recreational usage. One example of such a camera is the Microsoft Kinect sensor originally usedwith the Microsoft Xbox 360 game console. In this master thesis a system is presented that utilizes this device in order to create an as accurate as possible 3D reconstruction of an indoor environment. The major novelty of the presented system is the data structure based on signed distance fields and voxel octrees used to represent the observed environment. / Under de senaste åren har flera olika avståndskameror lanserats på konsumentmarkanden. Detta har skapat många intressanta applikationer både i professionella system samt för underhållningssyfte. Ett exempel på en sådan kamera är Microsoft Kinect som utvecklades för Microsofts spelkonsol Xbox 360. I detta examensarbete presenteras ett system som använder Kinect för att skapa en så exakt rekonstruktion i 3D av en innomhusmiljö som möjligt. Den främsta innovationen i systemet är en datastruktur baserad på signed distance fields (SDF) och octrees, vilket används för att representera den rekonstruerade miljön.

Zugl.: Karlsruhe, Univ., Diss., 2005

München, Techn. Univ., Diss., 2004.