A High Accuracy CNN for Breast Cancer Detection Using Mammography Images
Gunjan Jha, Anshul Jha, Eugene B John · 2024
Breast Cancer (BC) is one of the most prevalent cancers and second leading cause of mortality among women. Several radiographic imaging techniques, such as mammograms, computed tomography (CT), magnetic resonance imaging (MRI), histopathological imaging (HI), etc. have made it viable to diagnose BC at an early stage. Deep learning (DL) has emerged as an aid to radiologists and pathologists in the detection and prognosis of BC, handling large amount of radiographic and histopathological images efficiently and accurately. The primary motive of this research is to develop a high accuracy convolutional neural network (CNN) to detect and classify BC using mammography images from publicly available datasets. The convolutional neural network classifier model is used to distinguish the malignant and benign cells in breast images. The performance of the CNN model is measured in terms of accuracy, F-1 score, precision, recall and confusion matrix. The proposed CNN model achieved an inference accuracy of 99.18%.