Identification of Uterine Fibroids using Enhanced Image Mining Techniques: Bio-Inspired Xenogenetic Based Extreme Learning Neural Network Classification with Improved Fireflies Hausdorff Distance

Background: Uterine Fibroids are common non-cancerous cell growth in muscular wall of the uterus and these do not create any symptoms, but the size and location of these fibroids lead to several problems for women such as pain and heavy bleeding during the menstrual cramps. Many algorithms have been developed and each one has its own merits and demerits. However, none of the algorithms has reached 100% accuracy in identification of fibroids. Methods: The proposed system will identify fibroids by using the Magnetic Resonance Imaging (MRI) and Bio Inspired Xenogenetic based Extreme Learning Neural Networks (BIXELNN) classification algorithm. Initially, Adaptive Median Filtering technique (AMF) is applied to MRI image to remove the noise. Improved Multi- scale Segmentation is used to segment the images into different regions. From the segmented region, shape, intensity and texture features are extracted and then the Improved Fireflies algorithm with Hausdorff Distance (IFHD) is applied to optimize those features for classification process. Conclusion: Finally, the classification is performed by using BIXELNN, achieving high accuracy than other approaches, i.e., 97.3%.