Fast Automatic Segmentation of Polycystic Ovary in Ultrasound Images Using Improved Chan-Vase with Split-Bregman Optimization

摘要

Polycystic ovary syndrome (PCOS) is an endocrine disorder, characterized by the formation of many follicles in the ovary. This disorder seriously affects women's health, causing obesity, infertility, cardiovascular disease and diabetes. PCOS is diagnosed using ultrasound imaging, which gives important information on the number of follicles and their size. Manual detection of these follicles is laborious and error-prone. This paper presents an improved active contour without edge method which can identify small follicles. The method is further enhanced by Spilt-Bregman optimization method to reduce the computational time and accurately segment the images. The resultant system is tested on PCOS ultrasound images with promising results.