Page 48 - FiM_03_2016

Results
Repeatability and reproducibility of no-error measurements were excellent with %GP=100% for all gamma methods. 1%/1mm and local normalization is able to detect all type of errors (1%/1mm with global normalization is not able to detect the systematic shift of 2,5 mm), but it could overestimates some errors that have not clinical impact. In the table, we reported the results of sensitivity and specificity of PF to detect clinically relevant errors.
To date, lots of the enrolled patients are still in treatment. Anyway, it is possible to present some preliminary results. Ten patients ended the treatment; in particular the following pathologies were treated: one mediastinum lesion, 1 leaver lesion, 1 hip bone lesion, 1 lung lesion, 3 prostate lesions, 1 acetabulum lesion and 1 eye lesion. A total of 387 PTV DVH dose points (D2%, D95% and Dmean) and 320 OARs dose points were evaluated for 129 fractions, as reported in figure 1 and 2. Seven %DE were found to be higher than 3%: of them 5 were related to the eye’s lesion. The eye’s lesion was the smallest of the considered lesions with a 9.3cc volume for the PTV (average PTV volume 187cc): that could explain the high percentage variation even for small dosimetric variations. Worst results were found considering the OARs: 49 %DE (17%) were higher than 3%. This behavior was comparable with the one noticed in the pre-treatment analysis of the recalculated DVH, where the most high parameter variations were in presence of low dose values in large volume or high dose levels in small volume.
Tesi di Specialità
Figure 1 %DE of 387 PTV DVH dose points
(D2%, D95% and Dmean) evaluated for 129 fractions
Figure 2 %DE of 286 OAR DVH dose points evaluated for 129 fractions
40