Weighted variance based scale adaptive threshold for despeckling of medical ultrasound images using curvelets

1055 Words3 Pages

Medical ultrasonography is an ultrasound wave based imaging technique used to scan internal body structures and is non-invasive, portable, accurate and economical. Due to such characteristics, this imaging modality has gained high importance in diagnostic field of radiology. However, a major limitation of ultrasound images is the presence of locally correlated multiplicative speckle noise that degrades the image quality. The speckle noise is generated by interfering echoes of a transmitted waveform at the time of acquisition. The overlapping of sound echoes approaching with random time periods and amplitudes yields complicated interfering arrangement called speckle noise [1]. Speckle noise is likely to create uncertainty and may conceal vital clinical facts, thus diverting the opinion of radiologists. Therefore, there is a need of despeckling technique, which is capable to restore an image without affecting important features and texture of the original image so as to assist the radiologists in making a more accurate diagnostic decision.
The main focus in this work is to develop a versatile despeckling technique to improve the quality of medical ultrasound images for better clinical diagnosis. The proposed technique is a variant of the recently published technique by Sudha et al. [2].The innovative aspects of the proposed technique are three-fold: (a) used curvelets to efficiently separate edges and noise, (b) revised criteria for defining the weights of window model used to calculate the weighted variance by exploiting the intra-band dependencies present in the curvelet coefficients, and (c) devised a window tuner to help radiologists to control the degree of smoothing of output image.
The paper is organized as follows. A brie...

... middle of paper ...

...ne in curvelets than denoising using wavelets [28]. Hard thresholding has been used to modify curvelet coefficients [ 28-32].
The literature survey reveals that very little work has been reported for despeckling using curvelets [29]-[32]. None of these techniques [29]-[32], take into account the directional intra-band dependencies of curvelet coefficients for calculation of the threshold, which are very important to make the threshold adaptive to signal variation. Despeckling is always a trade-off between noise suppression and loss of information, something that experts are very concerned about. It is therefore attractive to keep as much of important information as possible. In this paper, an attempt has been made to design a scale adaptive threshold despeckling technique in the curvelet domain for recovering an image without losing the fine curved vital details.

More about Weighted variance based scale adaptive threshold for despeckling of medical ultrasound images using curvelets

Open Document