Page 57 - FiM_03_2016

Tesi di Specialità
Figure 2 PDF of the angular positioning error for CW and CCW rotation.
Figure 3 PDF of percentage MU delivery error per CA.
Influence of phantom geometry of QA results
Table 1 reveals that a change of the phantom geometry might result in different gamma passing rate, suggesting that comparing QA results obtained with different instruments might be troublesome. The mean value and the spread of the gamma passing rate shrinks towards 100% for all the virtual dosimeters moving towards more lax criteria.
Even if the three planar dosimeters show similar average results, the PR% values obtained for the single plans are not correlated. Conversely the three cylindrical dosimeters are mutually correlated. The degree of correlation between the PR% obtained in patient and in the different virtual dosimeters is given in Table 2. Clearly, planar geometry is not a good geometry to perform VAMT pretreatment QA: none of the PR% obtained with the three planar dosimeters correlate with results obtained in patient. On the contrary, PR% obtained with the three cylindrical dosimeters are well correlated with PR% obtained in patient and the cylindrical volume is found to be the most correlated of the three. Figure 5 and 6 plot these correlations.
Table 1 Mean values, standard deviations and existence intervals of gamma passing rate obtained with the different virtual dosiemters.
49