Radiation dose from volumetric helical perfusion CT of the thorax, abdomen or pelvis

Abstract

Purpose

To evaluate the radiation doses delivered during volumetric helical perfusion CT of the thorax, abdomen or pelvis.

Materials and methods

The dose-length product (DLP) and CT dose index (CTDIvol) were recorded and effective dose (E) determined for patients undergoing CT (4D adaptive spiral) for tumour evaluation. Image noise and contrast to noise (CNR) at peak enhancement were also assessed for quality.

Results

Forty two consecutive examinations were included: thorax (16), abdomen (10), pelvis (16). Z-axis coverage ranged from 11.4 to 15.7 cm. Mean DLP was 1288.8 mGy.cm (range: 648 to 2456 mGy.cm). Mean CTDIvol was 96.2 mGy (range: 32.3 to 169.4 mGy). Mean effective dose was 19.6 mSv (range: 12.3 mSv to 36.7 mSv). In comparison mean DLP and effective dose was 885.2 mGy.cm (range: 504 to 1633 mGy.cm) and 13.3 mSV (range: 7.8 to 24.5 mSv) respectively for the standard staging CT thorax, abdomen and pelvis. Mean tumour CNR at peak enhancement was 1.87.

Conclusion

The radiation dose imposed by perfusion CT was on average 1.5 times that of a CT thorax, abdomen and pelvis. The dose is not insubstantial, and must be balanced by the potential clinical utility of additional physiologic data. Further efforts towards dose reduction should be encouraged.

Appendix

Conversion factor for Dose Length Produce (DLP) to Effective dose (E)

The rationale for the use of the conversion factors for Dose Length Produce (DLP) to Effective dose (E) for the CTs acquired at 80 kV/100 kV in this study is explained further in this Appendix.

The conversion factor is dependent on the dose to the various specific radiosensitive organs as specified in ICRP 103[23]. The ImPACT calculator (pre-release version 1.01 [26]) calculates these organ doses from both the primary and scattered x-rays using Monte Carlo techniques. The calculator enables beam qualities specific to the acquisition (CT model and kV), and the section of the body irradiated to be entered into the calculation. The ImPACT calculator then multiplies the organ doses by tissue weighting factors as specified by ICRP 103 to give an effective dose.

The calculator is however designed for a conventional CT mode, where the dose in the craniocaudal direction is uniform, and not a ‘shuttle’ mode as in the case of the 4D adaptive spiral mode. Thus the potential non-uniformity of the acquisition dose was investigated initially to ensure that use of the calculator was appropriate. The measured normalized variation of dose with the 4D adaptive spiral (11.4 cm z-axis coverage) is shown in Fig. 2. The measured dose profile shown in Fig. 2 demonstrates a uniform distribution, and this indicates that substantial errors will not be introduced by assuming a uniform dose distribution in our calculations for the 4D adaptive spiral mode.

Fig. 2

Normalised absorbed dose profile for 4D adaptive spiral of the pelvis. This demonstrates a similar dose distribution to conventional spiral CT

The 4D adaptive spiral mode uses a primary beam of a relatively short irradiated length (12.4–14.5 cm). The organ doses for each type of scan, chest, abdomen, pelvis will depend on the exact location of the CT primary beam relative to the ICRP radiosensitive organs. As the position of the tumor, and scan location will vary from patient to patient, this variation was investigated for each of the body areas imaged (thorax, abdomen, pelvis).

Appropriate values for CT (Siemens Definition), body area (chest, abdomen, pelvis), kilovoltage (80,100,100 kV), mean CTDI (51.9, 110.5, 131.5 mGy) and irradiated scan lengths, defined as the value of DLP/CTDI (13.8, 14.5, 12.4 cm) were entered into the ImPACT dose calculator. A range of conversion factor values, E/DLP, were found for chest, abdomen and pelvis scans as the position of the centre of the scan was incremented by 1 cm along the body. The results are summarized in Table 5:

Table 5 The range of conversion factor values, E/DLP, for thorax, abdomen and pelvis regions as the position of the centre of the scan was incremented by 1 cm along the body

In addition to confirm that the use of the imPACT calculator was appropriate, measured organ doses were compared to organ doses calculated by the imPACT software (Fig. 3). To perform this, a number of calibrated thermo luminescent dosemeters (TLDs) were place in an anthromorphic phantom (male ART pelvis) at the following organ sites: colon, bladder and skin. The phantom was scanned using the standard pelvis volumetric helical perfusion CT protocol with the centre of the scan directly over the centre of the bladder. The comparison of calculated and measured organ doses are shown in the Table 6.

Fig. 3

The imPACT mathematical phantom is shown, together with the position and length of the pelvic scan used (Reproduced with kind permission of P. Shrimpton)

Table 6 Comparison of calculated imPACT and measured organ doses

The mean measured organ doses for colon and bladder were less than that for the imPACT calculator, by about 20%. This may be due to the exact positioning of the TLDs and the size of the anthromorphic phantom. These measurements indicate that the use of the imPACT calculator for the relevant scan parameters is unlikely to underestimate the organ doses, and hence effective dose and conversion factor E/DLP. The measured skin dose did not correlate to the calculated dose as it is the maximum skin dose, as opposed to the skin dose averaged over all the body as given by the calculator. An average skin dose, 11 mGy, estimated as the ratio of the skin in the primary beam to the skin area of the whole body (ICRP 89 [27]) was less than the dose given on the imPACT calculator, as this simple estimate does not include scattered radiation. The maximum skin dose does however indicate that the skin dose is well below a threshold dose for deterministic effects.

In summary the conversion factors used for the thorax, abdomen and pelvis were the average values 0.019, 0.018 and 0.012 mSv/mGy.cm (range 0.009–0.022) respectively, derived as shown in Table 5. A further uncertainty when determining organ doses and effective dose given by the imPACT dose calculator, is that the results are only valid for a standard mathematical phantom, representing a 70 kg person.