SPECKLE REDUCTION AND CONTRAST ENHANCEMENT OF ECHOCARDIOGRAMS VIA MULTISCALE NONLINEAR PROCESSING

This paper presents an algorithm for speckle reduction and contrast en hancement of echocardiographic images. Within a framework of multiscal e wavelet analysis, we apply wavelet shrinkage techniques to eliminate : noise while preserving the sharpness of salient features. In additio n, nonlinear processing of feature energy is carried out to enhance co ntrast within local structures and along object boundaries. We show th at the algorithm is capable of not only reducing speckle, but also enh ancing features of diagnostic importance, such as myocardial walls in two-dimensional echocardiograms obtained from the parasternal short-ax is view. Shrinkage of wavelet coefficients via soft thresholding: with in finer levels of scale is carried out on coefficients of logarithmic ally transformed echocardiograms, Enhancement of echocardiographic fea tures is accomplished via nonlinear stretching followed by hard thresh olding of wavelet coefficients within selected (midrange) spatial-freq uency levels of analysis. We formulate the denoising and enhancement p roblem, introduce a class of dyadic wavelets, and describe our impleme ntation of a dyadic wavelet transform. Our approach for speckle reduct ion and contrast enhancement was shown to be less affected by pseudo-G ibbs phenomena, We show experimentally that this technique produced su perior results both qualitatively and quantitatively when compared to results obtained from existing denoising methods alone. A study using a database of clinical echocardiographic images suggests that such den oising and enhancement may improve the overall consistency of expert o bservers to manually defined borders.