ULTRASOUND IMAGE SEGMENTATION FOR DETECTING FOLLICLE IN OVARIES USING MORPHOLOGICAL OPERATION AND EXTRACTION METHODS

Abstract

There are three different types of ovaries in women such as normal ovary, cyst ovary and Polycystic ovary syndrome (PCOS). The normal ovary consists of 8-10 follicle with size 2mm to 28mm. If the number of follicles more than 9 with size less than 9mm, it is classified as Polycystic ovary syndrome (PCOS). Follicles can be detected through ultrasound images. Ultrasound images produce noises resulting in segmentation images cannot be perform properly. Therefore, to detect the follicles in the ovary is by using morphological operations with feature extraction. In this paper, the method of detecting the follicles is described which involves segmentation of follicles from ultrasound images using morphological method consists of erosion, open and dilation. Therefore, the number of follicles is counted and can save time for the doctor who has spent a lot of time to manually counting the follicles.