Page 66 - ARNM-3-3

Advances in Radiotherapy
& Nuclear Medicine Software impact in Ho dosimetry
166

contouring on the CT. In many patients, tumor margins Both the tumor and liver VOIs were delineated using
were not clearly visualized on the non-contrast-enhanced manual contouring. The tumor VOI was delineated
CT acquired along with the SPECT. Therefore, tumor according to contrast-enhanced CT images or, when
segmentation, performed by nuclear medicine/radiology feasible, the non-contrast-enhanced CT from the
physicians, was accomplished by fusing the SPECT images SPECT/CT.
with a recent contrast-enhanced CT.
The tumor and liver mean doses were calculated for
2.2.2. Q-suite each patient and then compared with Hermia results. Dose

The previous SPECT/CT reconstruction was also coverage of the lesions was also evaluated at three different
provided to Q-Suite. However, in contrast to Hermes, this points using DVH: D , D , and D , where D represents
70
85
X
50
reconstruction had limited significance, as the software the minimum dose received by at least X% of the given
performed its own patient-specific calibration (Figure 2). volume.
Consequently, the key element in the new reconstruction The comparison between the two datasets involved both
was the CF (distinct from that used in Hermia). Bland-Altman analysis and Wilcoxon signed-rank tests to
To determine this CF at the time of the treatment, assess agreement and determine the statistical significance
single or multiple VOI were defined for calibration and of differences in mean absorbed doses, D , D , and D .
50
70
85
provided assumed radiation activity within those regions All statistical analyses were conducted using R software.
at the time of injection. For this study, the whole-liver VOI
was used as the CF, which is the method recommended by 3. Results
the Q-Suite supplier. This study included 14 patients who underwent a scout
SPECT counts are converted into activity per voxel procedure followed by treatment. However, one patient
using the patient-specific CF and then into dose per voxel was excluded due to ventilation-induced artifacts in the
by convolving the activity with a holmium DPK for liver dose map causing an inaccurate dose distribution, with
tissue. the highest concentration-voxel mistakenly appearing

A B C D

Figure 1. (A-D) Panel overview of the voxel-level dose calculation in Hermia Voxel Dosimetry for Ho microspheres (adapted from Hippeläinen et al. )
166
19

166
Figure 2. Workflow for Ho microspheres dose calculation at the voxel level using Q-suite

Volume 3 Issue 3 (2025) 58 doi: 10.36922/ARNM025220023