ESTRO 2020 Abstract book

S358 ESTRO 2020

eigenmaterial characterization outperforms all DECT- based estimation methods (eigenmaterial and a ED-Z decomposition) for both parameters. RMSE is reduced from 1.1 to 0.8% for the ED, and from 1.8 to 0.8% for the proton stopping power. With a similar analysis, figure 1b demonstrates that the MARS scanner outperforms DECT to estimate all elemental mass fractions. Figures 1c-d respectively show errors on the VNC ED and iodine concentration for each iodine vial scanned. The MARS scanner generally outperforms DECT to estimate both quantities, with the RMSE averaged over all vials respectively reduced from 0.4 to 0.2% and 0.3 to 0.1 mg/mL with the MARS scanner. Figure 2a shows the quality of VNC ED maps and iodine concentration for all iodine vials. Figure 2b shows that beam hardening effects can be mitigated with the MARS scanner, due to the presence of high energy bins.

OC-0585 Quantitative imaging performance of MARS spectral photon-counting CT for radiotherapy M. Simard 1 , R. Kumar Panta 2 , H. Prebble 3 , A.P.H. Butler 4 , P.H. Butler 5 , H. Bouchard 6 1 Université de Montréal, Department of Physics, Montreal, Canada ; 2 University of Otago, Centre for Bio- engineering, Christchurch, New Zealand ; 3 MARS Bioimaging Ltd., MARS collaboration, Christchurch, New Zealand ; 4 University of Otago, Department of Radiology, Christchurch, New Zealand ; 5 University of Canterbury, School of Physics and Astronomy, Christchurch, New Zealand ; 6 Université de Montréal, Department of Physics, Montréal, Canada Purpose or Objective The objective of this work is to evaluate the quantitative imaging performance of MARS spectral photon-counting CT (SPCCT) for radiotherapy applications. Specifically, an experimental comparison of the quantitative performance of dual-energy CT (DECT) and SPCCT to estimate iodine contents and physical properties of human-like materials relevant to radiotherapy is performed. In human-like materials, the accuracy of quantities relevant to photon therapy, proton therapy and Monte-Carlo simulations, such as the electron density (ED), proton stopping power and elemental mass fractions is evaluated. For iodine solutions, the accuracy of the iodine concentration as well as the virtual non-contrast (VNC) electron density are evaluated. Material and Methods Human tissue substitute phantoms, the Gammex 467 and 472, as well as 8 diluted iodine solutions are scanned with commercially available systems: the DECT (Siemens SOMATOM definition Flash dual-source) and the MARS SPCCT (MARS V5.0, MARS Bioimaging Ltd., Christchurch, NZ). Material characterization is performed in a maximum a posteriori framework with an optimized basis of materials tailored to characterize human-like materials and contrast agents in the context of noisy multi-energy CT data, based on the principles of eigentissue decomposition highlighted in previous studies (Lalonde et al. , Med. Phys. 44.10 (2017) and Simard et al. , Phys. Med. Biol. 64.11 (2019)). The resulting basis materials are called eigenmaterials. Results Figure 1a compares the root-mean-square error (RMSE) of the ED and proton stopping power averaged over all inserts of both Gammex phantoms. The MARS scanner with

Figure 1

Figure 2

Conclusion The MARS SPCCT can provide superior accuracy than a clinical DECT scanner to quantify physical parameters useful in radiotherapy. This highlights the potential of the technology for particle therapy, where more accurate