|
Activity Number:
|
544
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 2, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Physical and Engineering Sciences
|
| Abstract - #309902 |
|
Title:
|
Bayesian Wavelet-Based Despeckling of Ultrasound Medical Images Using the Gauss-Hermite Expansion
|
|
Author(s):
|
S. M. Mahbubur Rahman*+ and M. Omair Ahmad and M. N. S. Swamy
|
|
Companies:
|
Concordia University and Concordia University and Concordia University
|
|
Address:
|
Dept. of ECE, 1455 de Maisonneuve Blvd. W., Montreal, QC, H3G 1M8, Canada
|
|
Keywords:
|
Ultrasound image ; Despeckling ; Wavelet coefficients ; Gauss-Hermite ; Prior function ; Estimation
|
|
Abstract:
|
Wavelet-based despeckling of medical ultrasound images improves the diagnosis. Despeckling can be formulated as a statistical estimation technique, wherein the choice of an appropriate prior function of the sub-band wavelet coefficients is of major issue. In this paper, it is shown that a prior function that uses the Gauss-Hermite expansion performs better than standard ones, such as the generalized Gaussian or Bessel K-form, specifically because the proposed one can make use of an arbitrary number of higher order moments in parameter estimation. A Bayesian despeckling technique is then proposed in a homomorphic framework using the new prior function. Experiments are carried out using synthetically-speckled and real ultrasound images, and the results show that the proposed method performs better than several existing methods in terms of both the signal-to-noise ratio and visual quality.
|
