Comparative Analysis of Ovarian Images Classification for Identification of Cyst Using Ensemble Method Machine Learning Approach

Abstract

Generally, cysts are painless. When ovaries get enlarged there is a possibility for torsion and infertility. To detect the ovarian cyst, ultrasounds are used. In this work, ultrasound image dataset of ovaries of different women is been collected and compared with their accuracy with random forest, KNN, and ensemble method. Ultrasound images from the hospital as input from the system, then a pre-processing process is carried out to remove noise in the image. The next step is the segmentation results used for feature extraction by detecting cysts and their sizes. The proposed work is done by comparing the simple cyst and PCOS images for the classification of ovarian cyst along with the ensemble method. Random Forest, KNN, and ensemble methods are combined to compare the accuracy result of cyst types. By using this algorithm, a comparison graph is plotted.