138 research outputs found
An Iteration Method for X-Ray CT Reconstruction from Variable-Truncation Projection Data
No full textIn this paper, we investigate the in-situ X-ray CT reconstruction from occluded projection data. For each X-ray beam, we propose a method to determine whether it passes through a measured object by comparing the observed data before and after the measured object is placed. Therefore, we can obtain a prior knowledge of the object, that is some points belonging to the background, from the X-ray beam paths that do not pass through the object. We incorporate this prior knowledge into the sparse representation method for in-situ X-ray CT reconstruction from occluded projection data. In addition, the regularization parameter can be determined easily using the artifact severity estimation on the identified background points. Numerical experiments on simulated data with different noise levels are conducted to verify the effectiveness of the proposed method.</p
Segmentation of 2D and 3D Objects with Intrinsically Similarity Invariant Shape Regularisers
No full textThis paper presents a 2D and 3D variational segmentation approach based on a similarity invariant, i.e., translation, scaling, and rotation invariant shape regulariser. Indeed, shape moments of order up to 2 for shapes with limited symmetries can be combined to provide a shape normalisation for the group of similarities. In order to obtain a segmentation objective function, a two-means or two-local-means data term is added to it. Segmentation is then obtained by standard gradient descent on it. We demonstrate the capabilities of the approach on a series of experiments, of different complexity levels. We specifically target rat brain shapes in MR scans, where the setting is complex, because of bias field and complex anatomical structures. Our last experiments show that our approach is indeed capable of recovering brain shapes automatically.</p
Combined local and global image registration and its application to large-scale images in digital pathology
No full textA large-scale, nonlinear image registration problem can be partitioned into smaller independent subproblems by adding a global, coarsely discretized distance measure. The remaining inconsistencies between subdomains are smaller than without the coarse distance term and can be incorporated into the global solution by a blending method. Reaching a similar accuracy, the new method enables the registration of large-scale images that could otherwise not be computed.Ein großes, nichtlineares Bildregistrierungsproblem kann durch einen zusätzlichen globalen, grob diskretizierten Distanzterm in kleine, voneinander unabhängige Teilprobleme zerlegt werden. Die verbleibende Inkonsistenz zwischen den Lösungen der Teilprobleme ist kleiner als bei der Berechnung ohne den zusätzlichen Distanzterm. Die lokalen Lösungen können durch ein Mischungsverfahren zu einer globalen Lösung zusammengeführt werden. Das neue Verfahren ermöglicht die Registrierung von großen Bilddaten, die mit dem Referenzverfahren nicht berechnet werden können und erreicht dabei eine ähnliche Genauigkeit
A Fully Automated Approach to Segmentation and Registration of Medical Image Data for Pulmonary Diagnosis
No full textMolecular imaging is an exciting and relatively new technology that has found widespread use in the diagnosis and observation of various diseases. More recently, molecular imaging has penetrated areas such as drug development in order to facilitate the observation and analysis of the effects of newly developed drugs. The amounts of data in drug development experiments may be very large due to the fact that they contain both spatial and temporal information of medical images. Imaging techniques can facilitate the analysis of large amounts of data by automating information extraction and providing meaningful results. The focus of the project concerning this thesis is to create a emporal and spatial atlas of an animal by utilizing and integrating data from images of different modalities. More specifically, the application treated in the thesis makes use of ventilation and perfusion data from CT and SPECT scans in order to aid in the observation of the effects of newly developed drugs in the treatment of lung diseases. This thesis describes the segmentation and registration techniques used in detail and how these were utilized to align and combine ventilation and perfusion data from both CT and SPECT scans.Master of Applied Science (MASc
Combined local and global image registration and its application to large-scale images in digital pathology
No full text- …
