Abnormality Detection and Segmentation in Breast Digital Mammography Images Using Neural Network
Nixon Dutta, Basabi Chakraborty · 2021 11th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS) · 2021
Early diagnosis of breast cancer increases the patient's chances of survival. In the last two decades many Computer Aided Detection(CAD) systems were developed to help radiologist analyse digital mammograms. The current state-of-the-art systems use YOLO, Faster R-CNN and other classification and detection algorithms to detect masses and calcification present in the mammograms. The training process in these systems is complex, time consuming and requires large amount of data. The heterogeneous nature of the data with irregular shapes and sizes makes the task more difficult. In the present paper we illustrate the development of a deep learning study aimed to detect masses, calcification and other irregularities in digital mammography images using Anomaly Detection and Localization. The proposed model has been implemented using two publicly available datasets and satisfactory results have been achieved.