Not applicable, Dosimetric comparison, Radiation safety, Dosimetry, Diagnostic procedure, PET-CT, CT, Paediatric, Nuclear medicine, Hybrid Imaging, Multimodality Imaging
C. Cox, M. Segbers, M. van Straten; Rotterdam/NL
Findings and procedure details
LdCT scans were acquired on a 128 mCT scanner from cylindrical water/perspex phantoms with circumferences of 28, 31, 50, 63 and 100 cm (Fig. 1) using the clinical child and clinical adult protocols with an equal Quality reference CTDI (mGy). To understand the differences between these two protocols with respect to the AEC, the phantoms were also scanned with an adapted adult protocol based on the tube voltage, Quality reference mAs and Quality reference CTDI of the child protocol. The acquisition parameters used for the different protocols are listed in Table 1. The AEC strength was set to ‘strong’ for the child and to ‘average’ for slim and obese adults. This way, extra dose reduction for children was obtained while accepting a higher noise level for the anatomical correlation. For all phantoms, the CTDIvol (mGy) of the three protocols was registered to determine the influence of the AEC curve strength on dose. To compare the image quality between the protocols, image noise values were obtained by measuring the standard deviation of the Hounsfield numbers in a region of interest. All scans were repeated once and measurements were averaged for the analysis.
Fig. 2 plots effective load (mAs) versus circumference. As can be seen the effective load lower limit of the scanner was reached for the 28 and 31 cm phantoms for all protocols and for the 50 and 63 cm phantoms acquired with the clinical adult protocol. Furthermore, for the 50 cm phantom the mean effective load of the adapted adult protocol (22 mAs) is 46% higher compared with the clinical child protocol (15 mAs). It appears that the clinical adult protocol will provide the lowest mAs-value for all phantoms. However, mAs is not a direct measure for dose.
In order to compare dose differences between the protocols, CTDIvol is a better indicator (see Fig. 3). This figure shows that scanning with the clinical adult protocol, which reached the lower effective load limit of the scanner, results in a high mean CTDIvol in phantoms with circumferences until 63 cm. It can also be seen that the mean CTDIvol of the 50 cm phantom measured for the adapted adult protocol is higher compared to the clinical child protocol. This indicates that an adult based protocol uses the slim or obese adult AEC curves and never the child AEC curve.
Fig. 4 shows the mean image noise as a function of circumference. The trends were comparable for all protocols above a circumference of 63 cm. The sharper rise of the clinical child protocol curve above 63 cm might be caused by underexposure due to the use of the child bowtie filter. Further differences in image quality can be explained by differences in reconstructed slice thickness and field-of-view. The 50 cm phantom scanned with the clinical child protocol showed a high mean image noise compared to the adapted adult protocol (Fig. 5), which is also affected by the difference in AEC strength.