Ultrasound image despeckling based on MMSE estimation in nonsubsampled contourlet domain

To suppress speckle noise and preserve edges information in ultrasound images, the nonsubsampled contourlet transform (NSCT) is applied to decompose the ultrasound image into NSCT subbands. The multiplicative speckle noises in NSCT high frequency subbands can be expressed in additive forms. A thresholding method is applied to extract and preserve strong edge coefficients in each NSCT subband. Then a Bayesian minimum mean square error (MMSE) criterion based equation is achieved to despeckle other NSCT coefficients. Last the despeckled image is reconstructed by the inverse NSCT transformation. The experimental results of synthetic speckle and clinical ultrasound images show that the proposed method outperforms several ultrasound image despeckling methods in terms of speckle reduction and edge preservation indices.© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.