A novel approach for Breast Tumour Detection and Segmentation from MRI Images Using Image Processing Techniques

Abstract

Cancer's pre-stage, tumour, has grown to be a major issue in this day and age. Researchers are working to create solutions and cures for it. Breast tumours are remarkable cell enhancements in the brain tissue that are sometimes difficult to see with imaging techniques. A method used to show a detailed image of the affected breast area is magnetic resonance imaging (MRI). The medical imaging trick is an important behaviour in the disease's identification. In this study, a breast MRI image is selected for investigation, and a technique is intended to provide a clearer look of the area the tumour has attacked.. Anisotropic filtering for noise removal, bounding box classifier for segmentation, and morphological operations for differentiating the damaged area from the normal one are the important phases in the described technique, which takes as its input aberrant MRI breast images. The foundation of this procedure is obtaining sharp MRI pictures of the breast. The tumour is identified by categorising the pixel intensities in the filtered image. The results of the experiment demonstrated that the anisotropic filter had a 95.5% accuracy rate. Finally, for easy identification, the segmented region of the tumour is highlighted on the original image.