A Versatile Statistical Model for Despeckling of Medical Ultrasound Images

This paper presents a new despeckling technique for medical ultrasound (US) images based on a versatile statistical model. The method uses the generalized Gaussian distribution (GGD) and generalized gamma distribution (GGAD) to model the image and the speckle respectively in the detailed subbands of wavelet decomposition of log-transformed ultrasound image. Combining these a priori distributions with the Bayesian maximum a posteriori (MAP) criterion, shrinkage estimators are derived for processing the wavelet coefficients of the detail subbands. The visual comparison of despeckled US images and the higher values of quality metrics indicate that the new method suppresses the speckle noise well while preserving the texture and organ surfaces.