7. Relative electronic densities vs CT number using two different phantoms: treatment planning impact

Introduction This study succeed to a first published study [1] which shows HU variations for fix electronic densities with variation of CT acquisition parameters. Deviations up to 250 UH have been highlighted when a second phantom is used to make a new curve. We quantify this influence on several treatment plans. Methods Phantoms Catphan (Phantomlab) and 062M (CIRS) were used on a Lightspeed RT16 (GE) CT scan to establish the electronic densities/CT number curves. For each curve, 35 treatment plans were recalculated on Eclipse V11 (VARIAN) using AAA V10 (VARIAN) algorithm. For organs at risk (OAR) and CTV/PTV, values recorded are DAve (Average dose), DMin (Minimal dose), D98 (Dose gived at 98% of volume), D95, D80, D60, D40, D20, and DMax (Maximal dose). The relative dose deviations between the two plans, normalized to the prescribed dose, are presented. Results Whatever the technique (VMAT, SBRT, 3DCRT), the fractionation or the localization, the average deviations are in [−0.5; 0.5]%. By discriminating OAR and CTV/PTV, the average deviation for OAR group is (0.3 ± 0.4)% and (0.9 ± 1.3)% for CTV/PTV group. For the CTV/PTV group, higher variances appear for the sub group “pelvis/prostate/endometrium” and “brain” and are respectively 0.9% and 1.6%. The maximum average difference is 2.4% and the maximal difference is 3.5%. Conclusions Phantom choice seems to be an important parameter in the establishment of the electronic densities/CT number curve. Few studies [2] , [3] show this influence but a particular attention must be done for this curve in particular for anatomical regions of high electronic densities.