Project Director: Min Xian
Project Team: Aleksandar Vakanski
Breast cancer is one of the leading causes of death in females. Early detection of breast tumors is critical to increasing the survival of women diagnosed with this disease. This project serves as a key step to achieving the long-term goal of making accurate early detection of breast cancer available to more women at lower cost via less expensive and even portable devices. It focuses on improving existing approaches by modeling breast anatomy and implementing new deep learning architectures for breast tumor segmentation. If successful, the resulting models will accurately segment images of varying quality, collected with different imaging devices and settings. Breast ultrasound images from four medical schools will be used in this research. The success of the proposed project will improve the robustness and accuracy of tumor segmentation methodologies and will broaden the use of computer-aided diagnosis in the early detection of breast cancer in clinical practice.