Comparisons of hounsfield unit linearity between images reconstructed using an adaptive iterative dose reduction (AIDR) and a filter back-projection (FBP) techniques

Abstract

Background: The HU linearity is an essential parameter in a quantitative imaging and the treatment planning systems of radiotherapy. Objective: This study aims to evaluate the linearity of Hounsfield unit (HU) in applying the adaptive iterative dose reduction (AIDR) on CT scanner and its comparison to the filtered back-projection (FBP). Material and Methods: In this experimental phantom study, a TOS-phantom was scanned using a Toshiba Alexion 6 CT scanner. The images were reconstructed using the FBP and AIDR. Measurements of HU and noise values were performed on images of the “HU linearity” module of the TOS-phantom. The module had five embedded objects, i.e., air, polypropylene, nylon, acrylic, and Delrin. On each object, a circle area of 4.32 cm2 was drawn and used to measure HU and noise values. The R2 of the relation between mass densities vs. HU values was used to measure HU linearities at four different tube voltages. The Mann-Whitney U test was used to com-pare unpaired data and p-value < 0.05 was considered statistically significant. Results: The AIDR method produced a significant smaller image noise than the FBP method (p-value < 0.05). There were no significant differences in HU values of images reconstructed using FBP and AIDR methods (p-value > 0.05). The HU values acquired by the methods showed the same linearity marked by coinciding linear lines with the same R2 value (> 0.999). Conclusion: AIDR methods produce the HU linearity as FBP methods with a smaller image noise level.