Program Information
Simulation of Contrast Enhanced Digital Mammography with a Four Dimensional Digital Breast Phantom
Y Zhong*, C Lai, Y Shen, T Wang, C Shaw, UT MD Anderson Cancer Center, Houston, TX
WE-G-103-7 Wednesday 4:30PM - 6:00PM Room: 103Purpose: Although contrast injection helps make cancerous lesions more visible in contrast enhanced digital mammograms (CEDM), the overlapping of dense tissue with the lesion could affect the degree of the improvement. In this study, we developed and used a simulation method to investigate such effects.
Methods: A digital phantom for a compressed breast was generated from cone beam CT images of a mastectomy specimen to represent the stationary anatomy of a breast. The 3D map of a tumor was inserted into the phantom. Published contrast enhancement data and clinical contrast enhanced CT images were used to estimate and assign temporarily varying CT enhancement values to both cancerous and normal tissues following contrast injection. These values were added to the digital phantom to form a 4D digital phantom, which was used to compute digital mammograms to visualize the lesion and measure the time density curve over the lesion. The results were used to investigate and quantify the effects of the lesion size and overlapping with dense tissue structures on the visibility of the lesion, degree of contrast enhancement and time of arrival in CEDM.
Results: We have successfully generated and used a 4D digital breast phantom to simulate CEDM. It was found that the overlapping of the lesion with dense tissue could affect the degree of contrast enhancement and time of arrival information for the lesion. The effects was found to vary with the size of the lesion relative to the breast thickness in the direction of the x-ray path.
Conclusion: The benefit of contrast enhancement in digital mammography varies with the size of the lesion relative to the density tissue in the region of interest. The overlapping dense tissue structure could degrade the visibility of the contrast enhanced lesion both in the image and in the time density curves measured.
Funding Support, Disclosures, and Conflict of Interest: This work was supported in part by grants CA13852 and CA124585 from the NIH-NCI.
Contact Email: