Artificial Intelligence approaches on Ultrasound for Breast Cancer Diagnosis

Abstract

Among all the cancer types, breast cancer is the leading cause of death for women worldwide. Breast cancer affects women of all ages after puberty in every country on the planet, with rates rising as they become older. Early detection of breast cancer, before it spreads, allows for more effective therapies and hence improves survival rates dramatically. Breast cancer screening with mammography, ultrasound, and magnetic resonance imaging (MRI) can considerably improve a patient’s prognosis. Artificial Intelligence (AI) has consistently outperformed the competition in classification tasks a nd is currently being researched for use in breast cancer screening. Mammography is the most widely adopted imaging technique to check for breast cancer and early detection, but it has limitations and, like MRI technologies, it is not always accessible, especially in environments with limited resources. For these reasons, ultrasound has a significant impact on the diagnosis of breast cancer, both as an additional modality to mammography and MRI, and as a primary imaging modality in some areas. Breast ultrasound interpretation is a difficult task. Due to the difficulty of reading the images and the results, the use of breast ultrasound has been criticized for the increase in the number of false positives. In the present paper we refer to its applications in breast medical ultrasound imaging, such as lesion detection, segmentation, and classification, breast density evaluation, and breast cancer risk assessment; we refer to recent AI proposals aiming to predict the risk that a woman could develop breast cancer, also examining the limitations and future prospects of using AI in breast medical imaging.

Topics & Concepts