927 / 1096
S911
ESTRO 36
_______________________________________________________________________________________________
The patients were CT scanned with a 2
nd
generation dual
source CT scanner, SOMATOM Definition Flash (Siemens
Healthcare, Forchheim, Germany). DECT images were
acquired at 100/Sn140 kVp, and SECT images were
obtained as a weighted summation of the low and high
DECT images. The DECT scans were acquired at the same
day as the control CT scan midway through the treatment
course and using the same dose settings as used for the
control scan. The CT scans covered the whole anatomical
region of the head down to the top of lungs – the SPR
comparison was thereby performed over very
heterogeneous tissue regions. SPR images were calculated
from both the DECT and SECT scans for the four first
patients included in the study. For DECT, SPR images were
calculated using a noise-robust method previously
developed in our group. For SECT, the stoichiometric
method was used. Based on SPR images, difference maps
were calculated. Seven regions of interest (ROIs) were
placed, each covering a single tissue type. Relative SPR
differences between the DECT and SECT calculations were
extracted from the ROIs.
Results
For bone, SECT systematically underestimated the SPR
compared with DECT, while the reverse was the case for
the soft tissues (Fig. 1). The relative SPR differences
ranged from -2.2% to 0.9%, with a mean difference of -
0.6% (Fig. 2). Large variations of up to 1.5 percentage
points were seen for the SPR difference across the
patients. However, the differences for the individual
patients were systematically either positive or negative
for each ROI (Fig. 2).
Conclusion
Large differences in proton SPR estimation were found
between DECT and SECT, although these were within the
uncertainties which are currently used for dose
calculation in particle therapy. These differences indicate
that DECT will allow for reduction of treatment margins,
resulting in better dose conformity. We are currently
performing proton treatment planning for the patients
comparing DECT- and SECT-based proton SPRs to
investigate the dose difference in the tumour and in the
surrounding healthy tissues, as well as potential impact on
the range uncertainty margins used in proton treatment
planning.