Breast imaging is an important area of research with many new techniques being investigated to further reduce the morbidity and mortality of breast cancer through early detection. Computerized phantoms can provide an essential tool to quantitatively compare new imaging systems and techniques. Current phantoms used in breast imaging research lack sufficient realism in depicting the complex three-dimensional (3D) anatomy of the breast. Also, they are frequently limited to a single application, typically mammography, and do not have the flexibility to be applied to other emerging modalities such as tomosynthesis or CT. The four-dimensional (4D) NURBS-based cardiac-torso (NCAT) phantom, which provides a realistic and flexible model of the human anatomy and physiology, is widely used in imaging research. Previously limited to only a male anatomy, the NCAT was recently extended to include a detailed, whole-body anatomy for both a male and female subject. Despite this advancement, a current limitation to the phantom is that the female breast is modeled using only a simple outer surface and does not include any anatomical detail. As a result, the NCAT is severely limited in its application to breast imaging research where it may have a profound impact. The goal of this work is to create a series of detailed 3D computational breast phantoms capable of realistically simulating a wide range of anatomical variations in health and disease and with the flexibility to model different compression states of the breast for various imaging modalities and to incorporate them seamlessly into the 4D NCAT phantom for breast imaging research. UC Davis has one of the only laboratories in the world which has acquired over a hundred high-resolution breast CT datasets of normal subjects as well those with disease. The normal background anatomy of the phantoms will be constructed using state-of-the-art computer graphics techniques based on this unique data. Through an analysis of abnormal 3D imaging data, modeling techniques will be developed to simulate a range of typical abnormalities (microcalcifications and masses) indicative of breast cancer. Finite element methods will be developed and validated to simulate different compression states of the breast, enabling the phantoms to be applicable to various imaging modalities. This work will provide the necessary foundation to quantitatively evaluate and compare existing and emerging breast imaging devices and techniques. Unlike current, simplified breast models, the proposed phantoms, with the ability to simulate realistic, predictive patient imaging data from anatomically diverse subjects using different acquisition methods, would provide a more complete assessment of imaging techniques, not just in terms of simplified physical characteristics, but in terms of clinically relevant performance. As such, the phantoms will provide a unique and vital tool for breast imaging research.
The goal of this work is to create a series of detailed 3D computational breast phantoms with the flexibility to simulate a wide range of anatomical variations (normal and abnormal) and different levels of compression and to incorporate them seamlessly into the 4D NCAT phantom for multimodality breast imaging research.
