Multiscale anisotropy analysis of breast tissue subtypes from mammography and pathology

University Of Maine Orono, Orono ME

Investigators

Linked publications & trials

Abstract

Summary Half of breast cancer cases occur in women with no identifiable risk factors. Approximately 50% of women have dense breasts, which increases their risk 3-5 fold compared to those with entirely fatty breasts, but only a subset of them will develop breast cancer. One explanation lies in tumor microenvironment (TME) changes that promote a breast tumorigenic environment. The organization of stromal and glandular breast tissue subtypes plays a crucial role in influencing tumor invasion dynamics. However, there has been limited research on mammographic subtypes of dense tissue and their potential links to risk. In the original R15 funding period the PI and his team showed that they can segregate mammographic dense tissue into two textural subtypes: active dense tissue, which is structurally reorganizing and links to cancer dynamics, versus passive dense tissue. More recently, they found that tissue restructuring associated with early tumor onset may be detectable via computational mammography prior to radiological diagnosis. The original R15 grant helped show that the amounts and temporal change of mammographic dense tissue subtypes may influence cancer risk. In this renewal, a novel and powerful computational technique is proposed to complement the textural analysis for mammogram and also for histology whole slide imaging both proximal and distal to the tumor and for several tumor subtypes. The proposed research will advance the education and career preparation of undergraduates within Maine’s only biomedical engineering program, nurture a growing partnership among the University of Maine, Spectrum Healthcare Partners and Maine Medical Research Institute, introduce two UMaine faculty to biomedical research, and support the early career development of two assistant professors.