X-ray mammography is the standard for the detection of breast cancer. Mammography is not ideal due to its use of ionizing radiation, its superposition artifacts, its inability to differentiate malignant from benign lesions, and its inability to detect 8-22% of palpable breast cancers. Echo ultrasound imaging is used as an adjunct to mammography for the determination of simple cysts, but has not proven effective for screening. Therefore, in the USA, there are approximately 600,000 annual biopsies of benign breast abnormalities. Ultrasound Computed Tomography (UCT) quantitatively images the acoustic properties of tissue. We will construct a next generation, 3D UCT imager using a cylindrical array of piezoelectric elements, acting as both transmitters and receivers. This arrangement allows for the collection of a large number of 3D projections. We will reconstruct a 3D image using a new simultaneous algebraic reconstruction technique and a novel 3D ray-tracing algorithm. Several limitations of early 2D UCT imagers hindered their clinical acceptance. These include: 1) technical (computer) limitations of the 70's and 80's , 2) 3D refraction and diffraction phenomena that were only addressed in 2D, and 3) the overlap of malignant and benign ultrasound tissue properties. We will use current computational power to extend UCT to 3D and use advanced image analysis to allow for lesion diagnosis. Thus, the specific aims are: 1) to construct a prototype 3D UCT imager, 2) to develop 3D reconstruction algorithm based on geometric acoustics, and 3) to characterize the 3D UCT imager using phantoms and compare the results to the clinical requirements.
Thesaurus Terms: breast neoplasm /cancer diagnosis, computer simulation, computer system design /evaluation, diagnosis design /evaluation, image processing, mathematical model, tomography, ultrasound imaging /scanning electric field, image enhancement bioimaging /biomedical imaging, phantom model
