Enhancement of ultrasound images using modified anisotropic diffusion model in non-subsampled shearlet domain
Anterpreet Kaur Bedi, Ramesh Kumar Sunkaria, Deepti Mittal · 2017
Ultrasound imaging is an indisputable image modality for clinical purposes. Unfortunately, it comes with the inherent speckle noise which results in the degradation of the texture information in these images, thus making the diagnosis harder. This paper proposes a new approach for despeckling the ultrasound images combining the multiscale anisotropic diffusion model with the non-subsampled shearlet transform (NSST). The method involves the decomposition of images using the Non-Subsampled Laplacian pyramid, resulting in a low and high frequency sub images. Modified anisotropic diffusion method is further applied to the coarser component, whereas, the finer component is subjected to shearlet function, resulting in noisy coefficients, which are further subjected to thresholding. This multidimensional and multidirectional method enhances the visual characteristics of the ultrasound images with not just the removal of speckle noise, but also with the preservation of more edges, thus enhancing the images effectively. The performance of the proposed algorithm is assessed for real ultrasound images. Results show that the method excels over the earlier proposed techniques in terms of preservation of edges and structural similarities. Moreover, when analysed qualitatively, it is observed that there is a substantial removal of speckle noise, thus making the diagnosis easier for the radiologists.