3D Reconstruction of Anatomical Structures Using Interpolation Techniques and local Approaches

The reconstruction of the surface is the process by which a 3D object is reproduced from a collection of discrete values that sample the shape. These values are generally called point cloud. Commonly, the reconstruction methods are based on the fundamental properties of the point clouds, which are the density samples, noise, missing data, and outliers. We aim to reconstruct the surface of anatomical structures from medical images; Consider two main problems that are missing data and the presence of noise. We resolve the missing data by generating new samples from a set of contours based on Shape Morphing techniques. If we add noise to the previous problem, we must change the focus and therefore use an implicit rebuild method that is solid for the problems presented above. Finally, we combine part of the previous proposals to solve a specific problem, that occurs when we reconstruct medical images, when a contour is bifurcated into another. The methods are evaluated with the public database from medical images and compared with the standardized algorithms of state of the art and the Hausdorff distance is used to measure the perfance

• Autores: Arteaga Daza, Luis Felipe
• Fecha de publicación: 2019
• Institución: Universidad Nacional de Colombia
• Repositorio: Universidad Nacional de Colombia
• Idioma: spa
• OAI Identifier: oai:repositorio.unal.edu.co:unal/76727
• Acceso en línea: https://repositorio.unal.edu.co/handle/unal/76727; http://bdigital.unal.edu.co/73448/
• Palabra clave: 3 D reconstruction; Pint cloud; Shape morphing; Bifurcation; Local merging; 3D reconstrucción; Nube de puntos; Forma cambiante; Bifurcación
• Rights: openAccess
• License: Atribución-NoComercial 4.0 Internacional