Follicle Prediction for Polycystic Ovary Syndrome Diagnosis from Ovarian Ultrasound Images using CNN

polycystic ovarian syndrome (PCOS) is the most common hormone related disorder worldwide. PCOS is caused by numerous tiny fluid-filled sacs called follicles, which are produced by the ovaries. As expected, the ovaries did not always release the eggs. The detection of multiple follicles from ultrasonography (USG) scans is one of the most effective methods for early diagnosis of PCOS and formulating a treatment strategy for patients with this illness. In this study, a deep learning technique for PCOS follicle prediction was used to train and test 1924 ovary USG images. A deep convolution neural network (CNN) model with VGGNet16 was used to extract significant features and classify PCOS and non-PCOS ovaries. The performance of the VGGNet16 model was evaluated using four models with different parameter values. The experimental results reveal that VGGNet16 sequential model 4 provides maximum 99.4% for classification of PCOS and non-PCOS ovaries.