P2F-1 A Speckle Removal Filter Based on Anisotropic Wiener Filtering and the Rice Distribution

In this paper we will focus on the problem of removing speckle in 3D ultrasound (US) images while maintaining the structure information unaltered. It is intended as a preprocessing step for further image analysis modules. Optimum Wiener filter technique will be used but modified in two senses: first, Rice distribution will be used prior to applying the Wiener filter to correct for the bias which may show up for low SNR regions in the image; secondly, a boundary model will be embedded within the estimation framework in order to maintain the signal structure and to avoid blurring the image. Some experimental results are also presented giving rise to very promising outcomes