Evaluation of the Relationship Between Tumor Thickness and Relative Pixel Value in Dual Energy (DE) Imaging
F Jung1*, J Luce2, J Roeske1, (1) Loyola Univ Medical Center, Maywood, IL, (2) DePaul University, Chicago, ILSU-E-J-196 Sunday 3:00PM - 6:00PM Room: Exhibit Hall
Purpose: To determine the relationship between simulated tumor thickness and signal intensity in a dual energy (DE) subtraction image.
Methods: A phantom was constructed using cork (simulating lung), tissue equivalent material, and pork ribs. While keeping the total phantom thickness constant (16 cm), the thickness of cork was varied from 8-12 cm. Simulated tumors with thicknesses ranging from 1-5 cm were placed within inserts inside the cork. Ribs were placed on top of the phantom such that they overlapped the simulated tumor on a planar radiograph. Planar x-ray images were subsequently acquired using 60 kVp (4 mAs) and 120 kVp (1 mAs) for the various combinations of cork and simulated tumor thicknesses. DE subtraction was performed for each image set to provide rib suppression. Pixel intensity values of the simulated tumor on DE images were obtained and plotted vs. tumor thickness.
Results: This study yielded a linear relationship between the simulated tumor thickness and DE pixel value. The relative pixel value increased with increasing tumor thickness. Moreover, the slope was nearly independent of the thickness of cork. The effect of the cork was to shift the intercept, with larger thicknesses of cork resulting in a more negative intercept value.
Conclusion: The results of this study illustrate a linear relationship between DE signal intensity and simulated tumor thickness. By demonstrating this relationship, DE imaging may be used as a quantitative means to quickly and efficiently assess 3D tumor response in lung cancer patients. Using this technique, DE imaging may provide a means of screening for patients who are candidates for adaptive radiotherapy.
Funding Support, Disclosures, and Conflict of Interest: Project supported by a grant from Varian Medical Systems.