Contour-Based Identification of Multicystic Ovary Morphology Using us Images of Ovaries

Poly/multi-cystic ovary syndrome (PCOS) is a common endocrine disorder affecting women of reproductive age. Early and accurate diagnosis of PCOS is important to prevent its long-term health effects. PCOM is the structure of Polycystic are visually viewed by use of various medical imaging tools such as ultrasound images and MRI scans. which are helps to visualize the morphology features of follicles in the either sides of ovaris. In this paper, study about the segmentation of follicles present in the ovary and mark it as contours. Contour segmentation is a well-established technique in image processing and computer vision for segmenting objects with clear boundaries in an image. The proposed approach involves preprocessing the input image using edge-preserving filters such as a median filter, a Gaussian filter, and a bilateral filter used to preserve edges. In this study, enhance the various sized cysts present in the ovaries. Based on the Rotterdam criteria, size and count of cysts place significant diagnose process. By the use contour segmentation to highlight the boundaries of the ovary and segment the each and every regions inside the ovary. This paper investigates the accuracy of distinct edge preservation filters which are used for pre-processing the images. After preprocessed, contours are drawn to specify the follicles regions with the threshold >0.2.The US images of ovaries are used to examine the multiple cysts in the ovaries, These images used in this paper to examine contours segmentation with distinct filters such as Bilateral filter, Gaussian filter and Median filter. The results show that contour segmentation using a Gaussian filter achieved high segmentation accuracy with an similarity coefficient of 0.94.