Automated Spot Mammography for Improved Imaging of Dense Breasts
Abstract
We are developing an automated spot mammography technique to improve imaging of lesions within dense breast tissue. During the first year of this project, computer programs have been developed to identify large dense regions in digitized mammograms. The programs also determine the minimum sized rectangle that bounds a dense region. Prototype devices have been designed and built to: collimate the x-ray beam to the desired region, restrain the breast during changeover to a spot compression paddle, and manually translate the spot paddle to the desired location. Preliminary tests were also performed on a commercial full-field digital mammography (FFDM) system. Sophisticated exposure control and technique factor selection on the FFDM improved the penetration of simulated dense regions in a breast phantom. Combined spot collimation and compression were found to increase lesion conspicuity with the FFDM. Rather than pursue research with our laboratory system, we have decided to build an add-on system for the FFDM, which has superior imaging properties. Initial experience with the prototype devices has indicated the need for design improvements. These will be implemented during the second year. Finally another improvement, the addition of stereo mammography to the auto spot procedure will be investigated.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2000
- Accession Number
- ADA387901
Entities
People
- Mitchell M. Goodsitt
Organizations
- University of Michigan