Comparisons with other data
Comparisons with other data
The data calculated with PCXMC versions 1.2–1.5 have been earlier compared to the organ dose conversion factors calculated in NRPB by Jones and Wall (1985) and Hart et al. (1994b, 1996b) and were found to agree well. This agreement was to be expected, because also their data were calculated using the phantom models of Cristy (1980). Reasonable agreement of PCXMC results has also been found in many comparisons with other dose calculations and phantom models or dose measurements, e.g., Tapiovaara et al (1997), Schmidt et al (2000), Schultz et al (2003) and Helmrot et al (2007). The agreement with the NRPB data still exists for PCXMC 2.0 for most irradiation conditions. Small differences are evident in some irradiation conditions, because the composition and density of the phantom tissues have been changed and the phantoms have been modified from the earlier versions of the program. This is depicted in Fig. 4, which compares doses in some organs calculated with PCXMC versions 1.5.2 and 2.0 for a PA-direction photon irradiation of the head and neck. As can be expected from the differences in the phantom models, doses to the brain and thyroid are higher in the new version, whereas the dose in the muscle tissue is lower. The change in the oesophagus model and the change in the composition of active bone marrow also result in differences between these program versions.
Figures 5 and 6 compare organ dose data for two x-ray examinations, adult PA chest and paediatric AP abdomen, from PCXMC versions 1.2 and 2.0 and the data of NRPB. For the purpose of comparison, the NRPB data below have been renormalized to correspond to an air kerma (free-in-air) of 1 Gy. It is seen that in these examinations the conversion factors of most organs have not changed appreciably from the earlier versions of PCXMC. Exceptions to this are the active bone marrow, oesophagus and thyroid, where changes in the composition, modelling or surroundings, respectively, have been made. Similar changes can be expected also for other organs that have been modified from the earlier version, e.g. the breast of the 15-year-old phantom.
As was already noted above, the organ doses calculated are strictly valid only for the phantoms used for the calculation. To illustrate differences between different phantom models, we have compared organ dose conversion data calculated with PCXMC to the data of Schlattl et al (2007) who have calculated dose conversion factors for whole body external exposure of photons using voxel phantoms ‘Rex’ and ‘Regina’ that are expected to be adopted as the standard human models by the ICRP. Figure 7 shows their conversion coefficient from air kerma to effective dose as a function of photon energy for total body irradiation from the front (AP-direction) and back (PA-direction). The conversion factors of PCXMC have been calculated as an average of the effective doses of two adult phantoms which have been matched to the height and mass of Rex and Regina. The agreement between the data calculated with PCXMC and the data of Schlattl et al (2007) is remarkable in the AP irradiation case, and reasonable in the PA irradiation case.
Figure 8 shows the photon energy dependence of the conversion coefficient from air kerma to the dose in salivary glands. The data of Schlattl et al (2007) are given for both their phantoms, Rex and Regina, and the data calculated with PCXMC are obtained by hermaphrodite phantoms matched to the height and mass of Rex and Regina.
The largest difference between the dose conversion factors of Schlattl et al (2007) and PCXMC are in the doses in bones: the values calculated with PCXMC are about 50 % larger, depending on the photon energy. This is probably due to the difference in bone modelling. PCXMC uses the homogeneous bone-approximation whereas the bone model of Schlattl et al consider cortical bone and spongiosa separately and calculate the dose in the spongiosa only: the effect of the cortical part is to shield the spongiosa, and the dose in the spongiosa is reduced. Generally, the doses reported in Schlattl et al (2007) correspond to the size-matched phantom data of PCXMC to within about 20 %, and are sometimes lower and sometimes higher. For some organs the data (e.g., thyroid, skin, lungs, female liver and female thyroid) agree notably better, within about 5–10 % for energies above 20 keV. Typically, the doses in the two different-sized phantoms vary less in PCXMC than they do in the data of Schlattl et al. An example of typical organ data, the stomach, is shown in Fig 9.
The usability of the phantom size modification feature of PCXMC has been demonstrated in an extreme case by Smans et al (2008) who calculated doses in two premature babies with weights of 590 g and 1910 g. The differences between the results were explained by the differences in the phantom models and the difficulty to place an x-ray field similarly in them. The dose conversion values in that paper were made with PCXMC 1.5.2. We have repeated the calculations with PCXMC 2.0, and obtained essentially the same results as were reported in the paper of Smans et al (2008); in this case the differences between the two PCXMC versions are caused mainly by the changes in the densities and composition of the phantom tissues. The data of Smans et al. (2008) and the newly calculated results for the chest AP examination of the smaller phantom are shown in Table 5. In order to demonstrate the effect of field location in such comparisons, the table also includes PCXMC 2.0 data calculated with a slightly larger field and a downward displacement of the x-ray field by 0.5 cm.
Similar organ dose differences between computational and voxel phantoms have also been seen in the papers of Staton et al (2003), Lee et al (2006c) and Pazik et al (2007); dose differences of the same order are obtained also in doses of different voxel phantoms (Zankl et al. 2002, Schlattl et al. 2007).
|Organ or tissue||Smans et al (2008)7x5 cm2 field, voxel phantom(Gy/Gy)||PCXMC 1.5
7x5 cm2 field
7x5 cm2 field
7x6 cm2 field
0.5 cm lower
|Active bone marrow||0.13||0.11||0.15||0.18|