Temporal Subtraction of Digital Breast Tomosynthesis Images for Improved Mass Detection
Abstract
The original goal of this project was to combine two methods developed to increase the sensitivity of breast cancer imaging digital breast tomosynthesis (DBT) and temporal subtraction. The combined technique would allow for easier and earlier detection of breast cancer than with either technique alone. Although the original scope of the project outlined using computer simulations and physical phantoms to evaluate the feasibility of the method, an IRB approved initial investigation was performed using temporally acquired DBT images of human subjects. The results showed that structural discrepancies, due to differences in positioning and compression between temporally spaced acquisitions, caused slight anatomical variations between reconstructed slices that significantly reduced registration accuracy and subtraction. Although, the feasibility of combining the two methods using simulation and physical phantoms may have shown to be promising, the initial results using actual human subject data did not indicate a significant improvement in breast cancer detection. These results prompted the change in scope of the project to focus on developing a realistic computerized breast phantom (RCBP) to be used for breast imaging research. The RCBP can be used for simulations that study imaging system design, acquisition protocols, reconstruction algorithms, and image processing techniques. The phantom developed in this project will provide a vital tool to investigate current and emerging breast imaging methods and techniques with the ability to simulate realistic, predictive patient imaging data.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 2009
- Accession Number
- ADA519456
Entities
People
- Christina M. Li
- James T. Dobbins Iii
Organizations
- Duke University