L’apprentissage profond appliqué à la reconnaissance des anomalies mammaires

Abstract

Breast cancer is one of the most common malignant tumors in women, which seriously affect women’s physical and mental health and even threat to life. At present, mammography is an important criterion for doctors to diagnose breast cancer. However, due to the complex structure of mammogram images, it is relatively difficult for doctors to identify breast cancer features. At present, deep learning is the most mainstream image classification algorithm. We present the architecture of our work which will allow us to choose the hyperparameters of a convolutional neural network; the experiments of this application will be tested on the mammographic image database DDSM (The Digital Database for Screening Mammography). An optimized model for the classification of mammographic tissues according to their types (Normal / Abnormal) represents a diagnostic aid system applied to mammographic images. The proposed approach was evaluated on 6688 regions of interest extracted from mammographic tissues.