Multiscale geodesic active contours for ultrasound image segmentation using speckle reducing anisotropic diffusion

Image segmentation is a fundamental but undoubtedly challenging problem in many applications due to various imaging artifacts, e.g., noise, intensity inhomogeneity and low signal-to-noise ratio. This paper presents a multiscale framework for ultrasound image segmentation based on speckle reducing anisotropic diffusion (SRAD) and geodesic active contours (GAC). SRAD is an edge-sensitive diffusion tailored for speckled images, and it is adopted here to reduce speckle noise by constructing a multiscale representation for each image where the noise is gradually removed as the scale increases. Then multiscale geodesic active contours are applied along the scales in a coarse-to-fine manner to capture the object boundaries progressively. To avoid boundary leakages in low contrast images, traditional GAC model is modified by incorporating the boundary shape similarity between different scales as an additional constraint to guide the contour evolution. We compare the proposed model with two well-known segmentation methods to demonstrate its superiority. Experimental results in both synthetic and clinical ultrasound images validate the high accuracy and robustness of our approach, indicating its potential for practical applications in other imaging modalities.