Abstract
In this paper, we propose a breast cancer monitoring system to be used as a part of smart healthcare. The breast cancer is one of the big reasons of death of young and elderly women, yet women often hesitate to go to the doctor due to the social and psychological barrier. A smart solution based on the cloud can facilitate the women to overcome this barrier. In the proposed system, mammograms can be obtained from the machine, and processed and analyzed in the cloud. In the system, we use a local binary pattern (LBP) and a co-occurrence matrix from the region of interest (where there is a possible mass) of the mammogram image. Several features are extracted from the LBP and the matrix to describe the mass. A support vector machine is used as a classifier. Experimental results show that the proposed system can achieve more than 98% accuracy while discriminating between normal and mass, and more than 87% accuracy while discriminating between benign and malignant.