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.
Sahli, H, Alrefaya, M, Vanhamel, I & Dinh Nho, H 2009, A Nonlinear Probabilistic Curvature Motion Filter for Positron Emission Tomography Images. in Scale Space and Variational Methods in Computer Vision. Lecture Notes in Computer Science, Vol. 5567, Springer Verlag, pp. 212-223. <http://www.springerlink.com/content/tu800077226731u8/>
Sahli, H., Alrefaya, M., Vanhamel, I., & Dinh Nho, H. (2009). A Nonlinear Probabilistic Curvature Motion Filter for Positron Emission Tomography Images. In Scale Space and Variational Methods in Computer Vision (pp. 212-223). (Lecture Notes in Computer Science, Vol. 5567). Springer Verlag. http://www.springerlink.com/content/tu800077226731u8/
@inbook{aa36f481efc04aed982f81e27ed8b3f4,
title = "A Nonlinear Probabilistic Curvature Motion Filter for Positron Emission Tomography Images",
abstract = "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.",
author = "Hichem Sahli and Musa Alrefaya and Iris Vanhamel and {Dinh Nho}, Hao",
note = "X.C. Tai, K. M{\o}rken, M.Lysaker, K.-A. Lie",
year = "2009",
language = "English",
isbn = "978-3-642-02255-5",
series = "Lecture Notes in Computer Science, Vol. 5567",
publisher = "Springer Verlag",
pages = "212--223",
booktitle = "Scale Space and Variational Methods in Computer Vision",
address = "Germany",
}