Breast cancer is the second leading cause of death in women and, despite some questions, it is generally agreed that early detection of malignancies improves the chances of successful treatment. There is mounting evidence to suggests that tissue elasticity might be a means to detect and classify breast lesions because hardness is a property strongly associated with cancer. This rationale is underscore by the prominent place that physical examination has and continues to hold in breast cancer screening. The principle hypotheses which underpins Project I is that breast tissue elastic properties are sensitive and specific predictors of pathology. We will address five specific aims: 1) Advance our data acquisition techniques to (i) induce larger displacements in more mass of tissue, (ii) decrease the image acquisition time through improved pulse sequence designs, and (iii) acquire data using a continuous vibration mode of operation. 2) Advance our model-based image reconstruction methodology in collaboration with the Computational Core to include (i) tissue dampening, (ii) multi-valued property recovery, (iii) transient effects and (iv) overlapping zone optimization, first in two and then in three dimensions. 3) Conduct transient effects and (iv) overlapping zone optimization, first in two and then in three-dimensions. 3) Conduct extensive simulations and phantom imaging experiments under controlled conditions to identify optimal mechanical stimulation, image acquisition and property estimation options from those investigated in Aims #1 and #2. 4) Develop a clinical system to estimate the mechanical properties of both breasts in reasonable times, demonstrate its feasibility and refine procedural protocols on a limited number of volunteers. 5) Recover the mechanical properties of breast tissue on patients with normal and abnormal mammograms, first through a test phase of evaluation followed by a prospective validation study executed in collaboration with the Clinical Core. Project I pursues a synthesis of existing and novel ideas which will address many of the current problems associated with breast elasticity imaging. First, harmonic mechanical stimuli which produce compressional or dilatational displacements will be developed. Phase contrast measurements of harmonic displacement will also serve to reduce patient motion artifacts while the use of multi-spectral vibrational stimuli will afford the opportunity to account for dissipational effects. Second, model-based property reconstruction methods will be deployed to exploit the multi-dimensional displacement field gradients which can be accurately measured in three dimensions. Further, the volumetric subsurface displacement data provided by the proposed MR system provides tissue response observations that are more extensive than those typically available in the model-based image reconstruction context which is likely to translate into improved spatial and contrast resolutions relative to those which have been achieved with elasticity imaging to date.
Showing the most recent 10 out of 129 publications
