Polycystic Ovary Syndrome Detection Using Contextual Ensemble Network and ELMAN Neural Network with Green Anaconda Optimization

Polycystic Ovary Syndrome (PCOS) is a metabolic reproductive disorder characterized condition by an extended menstrual cycle. There are many methods currently in use, but they all have major limitations. The prediction rate, which takes longer due to factors like heterogeneity is one of the main aspects of PCOS that makes it difficult. Moreover, there was no correlation between the network’s generalization ability assessment and precise predictions. The ELMAN Neural Network has been used to identify PCOS in order to eliminate the aforementioned problems. The ovarian ultrasound image is pre-processed with Fast Local Laplacian Filter (FLLF) and Brightness Preserving Bi-Histogram Equalization. The Contextual Ensemble Network (CENET) is used in the segmentation process and the textural features are extracted using the Projective Integral (PI) and the color features are extracted using the Color Auto Correlogram (CAC). Finally, an Elman Network with a Green Anaconda Optimization (GAO) is employed for classification purposes to diagnose PCOS. According to the results of the experimental research, the proposed ELMAN network has an accuracy of 95%, 93% for precision, 92.5% for recall, 90% for specificity, F1-score is 91%. Thus, the CENET with ELMAN Neural Network for PCOS detection from ultrasound images was considerably simpler and more efficient.