Breast Cancer:Tumor Detection in Mammogram Images Using Modified AlexNet Deep Convolution Neural Network
Emmanuel Lawrence Omonigho, Micheal David, Achonu Adejo, Saliyu Aliyu
Abstract
The improvement of system accuracy is a key issue in the detection and classification of tumors in digital mammographic images. This affects how radiologists make accurate analysis in the diagnosis of breast cancer. The goal of this research is to use augmentation techniques to improve system classification accuracy on a large number of datasets. A popular deep convolutional neural network (DCNN) architecture known as AlexNet was modified and used to categorize mammography images into two classes of benign (normal) and malignant (abnormal) tumors. The results demonstrated an overall system accuracy of 95.70%. It indicates an improved performance over traditional approaches in breast cancer diagnosis.