Record Details
Comparison of object detection and image quality between conventional digital mammography and digital breast tomosynthesis through contrast measurements and phantom image evaluation
ScholarsArchive at Oregon State University
| Field | Value |
|---|---|
| Title | Comparison of object detection and image quality between conventional digital mammography and digital breast tomosynthesis through contrast measurements and phantom image evaluation |
| Names |
Leek, Angela E.
(creator) Higley, Kathryn A. (advisor) Chakrabarti, Kish (advisor) |
| Date Issued | 2014-03-17 (iso8601) |
| Note | Graduation date: 2014 |
| Abstract | Advancements in digital detector technology over the past decade have facilitated the transition of mammography image acquisition from film screen to full field digital applications. Nearly 90% of the mammography performed in the United States utilizes digital imaging technologies. The ability to capture the image information in an electronic format has allowed for three-dimensional image reconstruction techniques to be applied to digital mammography images by employing tomosynthesis acquisition techniques. The ability for digital breast tomosynthesis to represent the breast tissue in a three-dimensional format has significant applications in clinical mammography to overcome the inherent limitation of overlapping tissue in current 2D mammography applications. The quality control requirements for mammography systems are based from regulations initially implemented through the Mammography Quality Standards Act of 1992. These regulations have been modified as digital mammography has developed and currently require that imaging modalities other than film-screen adhere to the quality control program outlined by the manufacturer of the equipment. A test to measure the signal to noise ratio (SNR) and signal difference to noise ratio (SdNR) is routinely used in two-dimensional digital mammography quality control programs to assess the ability of the system to adequately represent objects within the breast tissue structures. The purpose of this study was to understand the differences in object detection between two-dimensional (2D) mammography and digital breast tomosynthesis (DBT) applications. This study begins by examining potential variations in SNR and SdNR measurements of 2D image acquisitions over increasing simulated breast thicknesses. These measurements were then compared to SNR and SdNR measurements obtained for both projection images and reconstruction slices generated in the DBT modality using the same simulated breast thickness parameters as those acquired in the 2D modality. Human reviewer scores of contrast detail phantom images acquired under corresponding parameters were used to determine whether any measured differences in SNR and SdNR values correlated with changes in the ability of a human observer to visually perceive known phantom image objects. As the thickness of the attenuating material increased, differences were noted in the SNR and SdNR measurements as well as the reviewer scores when 2D and DBT results were compared. This study found that the measured SdNR decreased as the BR material thicknesses increased for both the 2D and DBT modalities; however, DBT demonstrated a greater change than was noted for 2D. Reviewer scores also indicated a difference in the ability to visualize the smallest phantom image objects between the two modalities. The scope of this study does not provide for a clear correlation between the decrease in SdNR and reduced reviewer visualization of test objects. However, the noted differences between 2D and DBT in SNR/SdNR and reviewer scores support the need for consideration of object detection parameters for each modality separately. |
| Genre | Thesis/Dissertation |
| Access Condition | http://creativecommons.org/licenses/by-nc-nd/3.0/us/ |
| Topic | Mammography |
| Identifier | http://hdl.handle.net/1957/47124 |
