Positron Emission Tomography (PET) is an important nuclear medicine imaging technique which enhances the effectiveness of diagnosing many diseases. The raw-projection data, i.e. the sinogram, from which the PET is reconstructed, contains a very high level of Poisson noise. The latter complicates the PET image's interpretation which may lead to erroneous diagnoses. Suitable denoising techniques prior to reconstruction can significantly alleviate the problem. In this paper, we propose filtering the sinogram with a constraint curvature motion diffusion for which we compute the edge stopping function in terms of edge probability under the assumption of contamination by Poison noise. We demonstrate through simulations with images contaminated by Poisson noise that the performance of the proposed method substantially surpasses that of recently published methods, both visually and in terms of statistical measures. © 2009 Springer Berlin Heidelberg.
CITATION STYLE
Alrefaya, M., Sahli, H., Vanhamel, I., & Hao, D. N. (2009). A nonlinear probabilistic curvature motion filter for positron emission tomography images. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5567 LNCS, pp. 212–223). https://doi.org/10.1007/978-3-642-02256-2_18
Mendeley helps you to discover research relevant for your work.