Watershed Segmentation for Medical Ultrasound Images

This paper presents an efficient method for biomedical ultrasound image segmentation based on watershed transformation. It consists of four major stages. These stages are pre-processing, multiscale morphological gradient, watershed segmentation and finally region merging. The proposed scheme is tested using a set of medical ultrasound images. Experimental results show that our proposed method can produce accurate contours in medical ultrasound images.