Breast cancer is one of the most common cancers in women worldwide. The main indicators of an early breast cancer are micro calcifications (MCs) which are found during a mammography screening. They are not always restricted to malignancies as they appear also in healthy breasts. Besides divulging if the patient has cancer or no, their characteristics can say more about the survival rate, patient therapy, outcome and prognosis. However, since mammography is a projection image, it can hide and make MCS appearance to vary significantly. Thus discriminating between benign and malignant MCs, is in many cases difficult.
To assist radiologists, Computer Aided Detection and Diagnosis (CAD) systems have been developed in order to detect the location and boundaries of the region of interest (ROI) and label it as benign or malignant. During my master thesis, a CAD system with a low number of false negatives and false positive rate was conceived when analyzing 3D high resolution micro-Ct images. Given the evidence on the link between image features of MCs, malignancy and tumor type, the main objective of my research is to develop a CAD system for breast cancer by exploiting MCs individual properties. We hypothesize that Deep Learning (DL) methods enriched with hand- crafted features, will outperform each of these methods separately, will classify malignant MCs in different breast tumor types and discriminate between healthy subjects and subjects suffering from breast cancer.
Runtime: 2019 - 2020