Locating the Position of Breast Tumor in MRI Images Using Modified Back Propagation Neural Network

Abstract

newline Cancer is the most growing disease in the world. As we know that it is caused by an uncontrolled newlinedivision of abnormal cells. Breast cancer is a second type of cancer which causes death of Indian newlinewomen s. In India where the diagnosis is done in last stage, as resources are very limited and newlinelarge number women die per day. In our work, we investigate a new machine learning model for newlinelocating a Breast Tumor. We have used Modified Back Propagation Neural Network Model newline(MBPNN) in MRI images. This proposed algorithm, based on Back Propagation Neural newlineNetwork (BPNN). BPNN is supervised learning of data mining. In our work, we investigate a newlinenew machine learning model for locating a Breast Tumor. BPNN is supervised learning of data newlinemining. Our model is fast and more accurate in locating the tumor in Breast images. We have newlinecombined our model with Support Vector Machine (SVM) and Distance classifier algorithm and newlinecreated hybrid SVM and hybrid distance classifier model. To find out accuracy of our proposed newlinealgorithms we have compared our algorithm with other models such as basic Back Propagation newlineneural network model, hybrid SVM and hybrid Distance classifier. In our work we have created newlinetrained dataset by running 1012 MRI images. Preprocessed database of 813 MRI images was newlinecollected from different hospitals and MRI screening center newline