Rapid, High Resolution 3-D Ultrasound Tomography
Abstract
Ultrasonic transmission tomography holds out the hope of being a discriminating tool for breast cancer screening that is safe, comfortable, and inexpensive. From its inception, however, this imaging modality has been plagued by the problem of how to quickly and inexpensively obtain the data necessary for the tomographic reconstruction. The specific aim of this research is to determine how best to adapt a new microfabricated ultrasonic sensor (currently under development for defense applications) into a breast cancer screening tool. The sensor converts an acoustic wavefront into a modulated optical signal over an entire imaging plane. Using this device, it should be possible to obtain the data necessary for 3D imaging of a breast in a short time, without ionizing radiation, and without the need for compression of the breast. The first year focused on refinement of the sensor design and development of reconstruction algorithms. We have improved the sensor speed and robustness. We have implemented two different types of 2D reconstruction algorithms, based on filtered backpropagation and adjoint methods of solution. We have developed acquisition protocols that obviate the need for optical calibration, and have performed an acoustical calibration. We constructed an ultrasound tank with a rotation stage, and we acquired data from a 3D phantom. The modeling simulations of the data and the actual data from the phantom were in good agreement. We have explored techniques of physician-friendly display, and developed an initial approach for data display from patients.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 2001
- Accession Number
- ADA405231
Entities
People
- Jeffrey S. Kallman
Organizations
- Lawrence Livermore National Laboratory