Tomographic Mesh Generation for OSEM Reconstruction of SPECT Images

Yao Lu,Bo Yu,Levon Vogelsang,Andrzej Krol,Yuesheng Xu,Xiaofei Hu,David Feiglin
DOI: https://doi.org/10.1117/12.813866
2009-01-01
Abstract:To improve quality of OSEM SPECT reconstruction in the mesh domain, we implemented an adaptive mesh generation method that produces tomographic mesh consisting of triangular elements with size and density commensurate with geometric detail of the objects. Node density and element size change smoothly as a function of distance from the edges and edge curvature without creation of 'bad' elements. Tomographic performance of mesh-based OSEM reconstruction is controlled by the tomographic mesh structure, i.e. node density distribution, which in turn is ruled by the number of key points on the boundaries. A greedy algorithm is used to influence the distribution of nodes on the boundaries. The relationship between tomographic mesh properties and OSEM reconstruction quality has been investigated. We conclude that by selecting adequate number of key points, one can produce a tomographic mesh with lowest number of nodes that is sufficient to provide desired quality of reconstructed images, appropriate for the imaging system properties.
What problem does this paper attempt to address?