A Systematisation of Polycystic Ovary Syndrome Using Ultrasonography Image Follicle Screening

Polycystic Ovary Syndrome(PCOS) which affects teen girls and women. It consists of many follicles occurs in the ovaries. This paper is presented here to detect the PCOS based on ultrasound images. The first stage consists of preprocessing which employs the anisotropic diffusion filter and wavelet transform based enhancement. The second phase involves segmenting the image using edge detection and assigning labels to the different segments. The third stage is feature extraction which is used to discover local binary pattern, histogram calculation, feature vector generation, classifier training and testing. Dataset is collected from kaggle repository. The final stage is classification that has AdaBoost and Bagging can be effective for cyst detection by aggregating the predictions of multiple base classifiers. Here the best accuracy is calculated as 97% while using the ensemble technique.