A based Bayesian wavelet thresholding method to enhance nuclear imaging

  • Authors:

  • Nawrès Khlifa;Najla Gribaa;Imen Mbazaa;Kamel Hamruoni

  • Affiliations:

  • Research Unit of Signal Processing, Image Processing and Pattern Recognition, National Engineering School of Tunis, Tunis, Tunisia;Research Unit of Signal Processing, Image Processing and Pattern Recognition, National Engineering School of Tunis, Tunis, Tunisia;Research Unit of Signal Processing, Image Processing and Pattern Recognition, National Engineering School of Tunis, Tunis, Tunisia;Research Unit of Signal Processing, Image Processing and Pattern Recognition, National Engineering School of Tunis, Tunis, Tunisia

  • Venue:
  • Journal of Biomedical Imaging
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Nuclear images are very often used to study the functionality of some organs. Unfortunately, these images have bad contrast, a weak resolution, and present fluctuations due to the radioactivity disintegration. To enhance their quality, physicians have to increase the quantity of the injected radioactive material and the acquisition time. In this paper, we propose an alternative solution. It consists in a software framework that enhances nuclear image quality and reduces statistical fluctuations. Since these images are modeled as the realization of a Poisson process, we propose a new framework that performs variance stabilizing of the Poisson process before applying an adapted Bayesian wavelet shrinkage. The proposed method has been applied on real images, and it has proved its performance.