An Automatic Detection of Breast Cancer using Efficient Feature Extraction and Optimized Classifier Model

Image Processing Techniques (IPTs) are widely used in healthcare. IPTs work on images to extract useful information from them. IPTs are effective in the early diagnosis of cancers which helps in their proper treatment. The number of human beings affected by cancer has drastically increased recently and more in women in the form of Breast Cancers (BCs). This work is on detecting BCs from image datasets using IPTs. This work proposes the use of Weiner Filters (WFs) in Preprocessing BC image data. WFs are applied to images as they can improve the quality of images. Image segmentations are executed using the Watershed Algorithm which detects Regions of Interest (ROIs) in images. Feature Extractions are done using the statistical method Gray Level Co-occurrence Matrix (GLCM). These extracted features are then selected using Black Widow Optimization (BWO) Algorithm. The selected features are then trained by Fuzzy Neural Networks (FNNs) which classify BC cases. The results of the classifier when evaluated were found to have a better performance on medical images in terms of Accuracy.