Abstract

Cancer incidence and mortality have been increasing at an accelerated pace over the past 3 decades globally, making cancer the major public health problem. Among females, breast cancer is known as the most diagnosed cancer and the main cause of cancer deaths in more than 100 countries. In 2018, there are about 2.1 million newly diagnosed breast cancer cases around the world, responsible for nearly 1 in 4 cancer cases among females. However, the causes of breast cancer are still not clearly known to doctors. Early diagnosis of breast cancer can make the disease easier to treat. Several diagnosis techniques are commonly used to distinguish malignant breast tumors from benign ones. Fine Needle Aspiration (FNA) is a well-known procedure used to diagnose breast cancer, but it suffers from a lack of satisfactory diagnosis performance. For FNA, radiologist, oncologist, and pathologist are required to render final judgment together in breast cancer diagnosis, which is time-consuming. Also, there is higher possibility to give rise to errors due to exhaustion or inexperience, which panic patients when false-positive result happens or miss optimum treatment time when false-negative result appears. Therefore, developing an efficient diagnosis support system to assist doctors’ diagnosis of cancer has great significance for medical diagnosis process

Description