Quantitative Evaluation of Relative Electron Density to HU Calibration Stability In Ethos Hypersight CBCT for Adaptive Radiotherapy
Abstract
Purpose
To characterize the dependence of relative electron density (RED) to Hounsfield unit (HU) calibration accuracy on acquisition and reconstruction settings on the Varian Ethos HyperSight CBCT to assess implications for adaptive radiotherapy dose calculation.
Methods
A RED phantom containing eight inserts was imaged using the Ethos HyperSight CBCT using clinically relevant acquisition modes at 100 (head), 125 (pelvis), and 140 kV (large pelvis). Reconstruction algorithms included filtered back-projection (FDK), iterative CBCT (iCBCT), iCBCT Acuros, and iCBCT with metal artifact reduction (iCBCT-MAR). Tube current was also examined at low, middle, and high mAs settings. Mean HU and HU standard deviation (SD) were measured for each insert, and the dependency on tube potential, reconstruction algorithm, mAs, and material was examined.
Results
All acquisitions demonstrated the expected bi-linear RED-HU relationship. Tube potential minimally influenced low-density materials, but did induce systematic HU scaling in higher-density inserts. For cortical bone, the mean HU exhibited a range of 333 HU between 100 and 140 kV. FDK and iCBCT yielded comparable RED–HU relationships, as did iCBCT Acuros and iCBCT-MAR, but a maximum difference of 419 HU was found between them for cortical bone. RED-HU relationships were unchanged with increasing mAs, but as expected, HU SD decreased with increasing mAs, plateauing beyond the middle setting. HU SD was dominantly material-dependent with limited dependency on reconstruction algorithm and tube voltage beyond 100 kV.
Conclusion
RED-HU calibration on the Ethos HyperSight CBCT is influenced by tube potential, and reconstruction algorithm, with differences of up to 419 HU. This is an important consideration for Ethos, as the CBCT is used in adaptive dose calculation and a single calibration curve is used across all imaging modes. Our further research is examining the effect of these HU changes on dose calculation and the impact of field-of-view, phantom location, and phantom diameter.