Authors : Muayad S. Croock, Ayad E. Korial, Tara F. Kareem, Qusay Sh. Hamad and Ghaidaa M. Abdulsaheb
Abstract: Recently, the serious diseases that can attach people have been increased in scary way. One of these diseases is the breast cancer. This type of diseases affects the women in particular more than men. In this study, a deep-learning framework is proposed for detecting the breast cancer in early stages based on mammography images. It adopts the incremental development software process model as a part of software engineering in the designing framework for more reliability and extendibility. The presented method has been established on extracting the features of the employed images as a learning dataset for convolutional neural network inside the deep-learning strategies. Different step algorithms have been used for performing the detection of the standard benchmarks of breast cancer in the soft tissue, shown in the utilized mammography images. A real dataset, collected from Baghdad hospital is considered and it is divided into 30% test and 70% training sets. The obtained results show a high accuracy in terms of feature extraction of training set about 100% and breast cancer detection from test set as a validation accuracy about 90%.
Muayad S. Croock, Ayad E. Korial, Tara F. Kareem, Qusay Sh. Hamad and Ghaidaa M. Abdulsaheb, 2019. Software Engineering Model Based Early Detection Method of Breast Cancer using Deep-Learning Framework. Journal of Engineering and Applied Sciences, 14: 5775-5781.