Effects on surrounding tissue radiographic features in diagnosis of breast cancer
Abstract/Contents
- Abstract
- Breast cancer is the most deadly cancer for women in developing countries and the second most deadly cancer for women in developed nations. Mammograms are an essential component to early detection of breast cancer; however, only 20% of biopsied breast lesions are cancerous. Researchers have spent the last two decades developing computer-aided diagnosis (CADx) systems to aid in the screening and diagnosis process, yet no CADx system has progressed to clinical use. I believe that the research in CADx of breast cancer in mammography could be improved by incorporating imaging features from the tissue surrounding masses. It has been shown that breast density is related to cancer risk, and the environment of a mass seen on a cellular level can aid in predicting survival. This knowledge has not yet been utilized in our design of CADx systems. I hypothesize that the utilization of surrounding breast tissue features will improve the performance of CADx systems in mammography. In order to test my hypothesis, I have proceeded through three steps. First, I have developed a standard data set for use in CADx system evaluation. Second, I have developed methods of unsupervised feature learning for quantitative analysis. Finally, I have tested the developed feature learning methods on the mass and on the tissue surrounding a mass. The CADx system I built including the features from surrounding tissue achieved an A_Z of 0.80. This result exceeded that of all other methods tested on this data set. As such, the result supports my hypothesis that radiographic imaging features from the tissue surrounding masses are discriminatory for diagnosis and can be used to improve the performance of CADx systems.
Description
| Type of resource | text |
|---|---|
| Form | electronic; electronic resource; remote |
| Extent | 1 online resource. |
| Publication date | 2017 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Associated with | Lee, Rebecca Sawyer |
|---|---|
| Associated with | Stanford University, Department of Biomedical Informatics. |
| Primary advisor | Rubin, Daniel |
| Thesis advisor | Rubin, Daniel |
| Thesis advisor | Lipson, Jafi |
| Thesis advisor | Musen, Mark A |
| Thesis advisor | Napel, Sandy |
| Advisor | Lipson, Jafi |
| Advisor | Musen, Mark A |
| Advisor | Napel, Sandy |
Subjects
| Genre | Theses |
|---|
Bibliographic information
| Statement of responsibility | Rebecca Sawyer Lee. |
|---|---|
| Note | Submitted to the Department of Biomedical Informatics. |
| Thesis | Thesis (Ph.D.)--Stanford University, 2017. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2017 by Rebecca Lynne Lee
Also listed in
Loading usage metrics...