Generation of realistic breast models using neural networks for dosimetry in X-ray breast imaging techniques

Abstract

Breast cancer is the type of cancer with the highest incidence among women worldwide. The most effective method to reduce mortality from this disease is early detection, diagnosis and treatment. For early detection purpouse, mammography is the most recommended imaging technique in several contries for breast cancer screening. Besides, new imaging modalities have been proposed (such as tomosynthesis and breast tomography). Periodic exposure to ionizing radiation, related to mammography screening examination or different breast imaging modalities, may be related to a risk of carcinogenesis, since the breast is a radiosensitive organ. This risk can be estimated from the dosimetric quantity named Mean Glandular Dose (MDG), since the glandular tissue is the most prone to radiation-induced mutations. Despite the importance of this dosimetric quantity, it is usually estimated based on simplified breast modesl, which consideres the breast as a homogeneous distribution of glandular/adipose tissue. Thus, the application of DGM for the determination of reference dose levels in a specific population is limited. This project proposes the development of a methodology for breast dosimetry evaluation in different imaging modalities (mammography, breast tomosynthesis and breast CT), based on Monte Carlo simulations. Realistic breast models will be generated using neural networks for tissue classification from mammographic images. The implementation of the heterogeneous models in the simulations allows a more detailed study of the spatial dose distribution in the breast tissues, and opens the opportunity for dosimetric assessment at the patient-specific level. Thus, this project development will contribute to the strengthening of the research area of the Medical Radiological Physics Group (GFRMd)in the Department of Applied Physics of the Institute of Physics Gleb Wataghin. In addition, it will be able to contribute to the training of human resources in the field of Medical Physics. (AU)