Abstract: Ultrasound imaging is a widely used imaging modality for diagnostic purposes, due to its many advantages. However, the usefulness of Ultrasound imaging is degraded by the presence of signal dependent noise known as speckle. This paper proposes a model to segment the abnormality present in the Ultrasound image such as tumor or the lesion or the calculi. First an SRAD (Speckle Reducing Anisotropic Diffusion) filter is applied to the Ultrasound image to reduce the Speckle noise. This is followed by segmenting the abnormality by using active contour approach. Three methods of active contour segmentation are used. First, an edge based active contour called Geodesic active contour is applied. Next, the region based methods which are Chan Vese active contour model and a modified version of Chan Vese method are applied. The algorithm is used to segment different types of abnormality in images of different parts of the body like liver, pancreas, kidneys, breast, prostate and uterus. Segmentation results are compared and visually it can be observed that modified Chan Vese active contour method works better than Geodesic and Chan Vese method in segmenting the abnormality.

Keywords: Ultrasound images, Speckle noise, active contour segmentation, abnormality.