Task-Specific Optimization of Mammographic Systems
Abstract
This study sought to understand how different imaging parameters affect clinical diagnosis. First it developed research tools for measurement and simulation of mammographic imaging. Second we applied these research tools and conducted a large human observer experiment to answer several clinically relevant questions. The first question explored the impact of display resolution on the detection of breast masses and calcifications. We found that different displays had little impact on clinical performance. The second question explored the effect of reduced dose on the detection of breast lesions. We found that the increased noise from reduced dose did impact radiologist performance. Reducing the dose by half did not have a statistically significant impact on diagnostic accuracy suggesting that mammographic dose could be reduced modestly with little impact on clinical performance. These results have immediate implications for clinical breast imaging.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 2007
- Accession Number
- ADA469423
Entities
People
- Robert N Saunders
Organizations
- Duke University