Regresar

Transfer Learning in Breast Mammogram Abnormalities Classification with Mobilenet and Nasnet

Abstract:

Breast cancer has an important incidence in women mortality worldwide. Currently, mammography is considered the gold standard for breast abnormalities screening examinations since it aids in the early detection and diagnosis of the illness. However, both identification of mass lesions and its malignancy classification is a challenging problem for artificial intelligence. Research has turned to the use of deep learning models in mammography which can enhance the performance of Computer Aided Diagnosis Systems (CADx). In this paper, we present our preliminary results on the use of transfer learning for malignancy classification of breast abnormality. We experiment with models that, according to our literature review, have not yet been explored thoroughly such as NasNet and MobileNet. Their performance is compared with InceptionV3 and Resnet50. The best results were obtained with Resnet50 and MobileNet with 78.4% and 74.3%, respectively. Also, some image pre-processing steps are studied in order to increase classification accuracy.