Cell-based region competition for ultrasound image segmentation

A new approach, cell-based region competition, based on watershed transform and region competition is proposed to detect the multiple objects of interest with highly winding contours in the ultrasound images. Multi-scale Gaussian filters are used to remove the speckle in the ultrasound image and preserve the edge information. Sobel filter are then applied to generate a gradient map. Immersion simulation of watershed transform is used to generate initial cells on the gradient map. Cell-based region merging and splitting are performed on the original ultrasound images based on the likelihood ratio tests of speckle noises. Two types of images have been used to evaluate the performance of the proposed cell-based region competition scheme, namely, the simulated phantom images with the speckle and the clinical ultrasound images. The experimental results on both types of images demonstrate the practical feasibility of the proposed cell-based region competition scheme for ultrasound image segmentation.