Can an ANN based radar data processing approach be an aid in breast cancer detection?

摘要

In this paper an Artificial Neural Network (ANN) based radar data processing approach has been assessed as a possible aid for the breast cancer diagnosis. The aim is not to achieve the imaging of the cancerous breast but to detect the existence or not of a tumor, independently by his depth and width. In the context of such purpose we use 2D realistic geometries of the healthy breast derived from the models provided by the UWCEM database. From simulated radar signals received around the breast we extract suitable preprocessed data for training and testing the used ANNs. The best ANN diagnosis performances have been obtained in the case of breast geometries for which we have assumed that the tumor isn't positioned behind the fibroglandular tissue.