Track: Modeling and Simulation

Abstract

Polycystic ovary syndrome is a common medical problem affecting women worldwide. It is attributed to excessive hormone production that causes an increase in the number of follicles in the ovaries. To detect this condition, specialist doctors examine ultrasound images by calculating the number and size of follicles in the ovaries. However, this procedure can yield inaccurate findings and endanger patients. Another tool used in this field is an ultrasonogram, but it is considered inaccurate because the noise from application processing can produce blurred images that appear to be clear contact segments. In this study, a segmentation method for detecting follicles in the ovaries is proposed to analyse the image quality of an ovarian segment. The watershed method and a combination of the edge enhancement method and Kirsch’s template are implemented to create and efficiently examine a segmented image of an ovary. Results show that the proposed method with the addition of a histogram on the threshold image can be effectively used to extract the morphological closing image of follicles. Follicles can be detected more clearly, and segmentation can be performed more efficiently. The mean square error and the peak signal-to-noise ratio of the combined methods and the watershed method have no significant differences. Thus, the combined method can produce clear images and reveal the number of follicles more quickly and easily than the watershed method.