Relationship between the radiation doses at nonenhanced CT studies using different tube voltages and automatic tube current modulation during anthropomorphic phantoms of young children

Abstract To compare the radiation dose and image noise of nonenhanced CT scans performed at 80, 100, and 120 kVp with tube current modulation (TCM) we used anthropomorphic phantoms of newborn, 1‐year‐old, and 5‐year‐old children. The noise index was set at 12. The image noise in the center of the phantoms at the level of the chest and abdomen was measured within a circumscribed region of interest. We measured the doses in individual tissues or organs with radio‐photoluminescence glass dosimeters for each phantom. Various tissues or organs were assigned and the radiation dose was calculated based on the international commission on radiological protection definition. With TCM the respective radiation dose at tube voltages of 80, 100, and 120 was 29.71, 31.60, and 33.79 mGy for the newborn, 32.00, 36.79, and 39.48 mGy for the 1‐year‐old, and 32.78, 38.11, and 40.85 mGy for the 5‐year‐old phantom. There were no significant differences in the radiation dose among the tube voltages and phantoms (P > 0.05). Our comparison of the radiation dose using anthropomorphic phantoms of young children showed that the radiation dose of nonenhanced CT performed at different tube voltages with TCM was not significantly different.

By using low tube voltage scans, decreasing the tube voltage increases contrast resolution, and the image noise may be higher at the same dose level. [7][8][9] For CT angiography studies, iodinated contrast material is used to improve enhancement without a substantial increase in the image noise. 9 Kalender et al. 10 recommended that lower tube voltages be applied at contrast-enhanced pediatric CT.
However, there are few reports on the radiation dose in pediatric unenhanced CT studies using lower tube voltages.
We used anthropomorphic phantoms of young children to evaluate the radiation dose for unenhanced pediatric scans performed at different tube voltages with tube current modulation (TCM).

2.B | CT Scanning
All CT scans were performed on a 64 detector row scanner (Lightspeed VCT; GE Healthcare, Milwaukee, WI) from the thorax to the lower abdomen including the entire lungs. The scan range was 265 mm for the newborn, 300 mm for the 1-year-old, and 405 mm for the 5-year-old phantom. The scanning parameters were helical mode, beam width 40.0 mm, section thickness 5.0 mm, pitch factor 0.984 mm/rotation, gantry rotation time 0.4 sec, 64 9 0.625 mm detector collimation, settings for the small scan field of view were 100, 150, and 150 mm for the newborn, the 1-year-old, and the 5- year-old phantom, respectively, matrix size 512 9 512, reconstructed mode for plus mode, standard reconstruction kernel. The applied tube voltages were 80, 100, and 120 kVp, and the tube current was changed from 20 to 330 mA to maintain the image quality (noise index at 12) using automatic tube current modulation for all phantoms (Figs. 3a-3c). 11 We verified that none of the scans reached the maximum mA. CT images were acquired with filtered back projection (FBP) algorithms under the standard kernel/filter.

2.C | Image noise and dose-length product at different tube voltages
At each tube voltage and in all phantoms we measured the image noise [standard deviation (SD) of the CT number] in the center of the phantom at the level of the chest and abdomen within a circumscribed 6.0 mm diameter region of interest (ROI). For each scan we acquired 100 consecutive images in the z direction using a CT workstation (Advantage Windows 4.4, GE Healthcare) to separate the 100 images from the upper chest to below the pelvis. The slice thickness was 5.0 mm for FBP algorithms using standard kernel/filter. The mean value for the SD of the CT number was calculated.
Dose-length product (DLP) values displayed on the CT console were recorded.

2.D | Dosimeters
Glass dosimeters are accumulation-type, solid-state dosimeters. 12 They take advantage of the radio-photoluminescence of silver-activated phosphate glass and are comprised of rod-shaped silver-activated phosphate glass, a plastic capsule, and an automatic reader unit. Their weight composition is P (31.55%), O (51.16%), Al (6.12%), Na (11.0%), and Ag (0.17%) and their effective atomic number and density are 12.039 and 2.61 g/cm 2 , respectively. Measurable doses range from 10 lGy to 10 Gy in standard mode and from 1 to 500 Gy in high-dose mode. We used GD352M glass dosimeters (Asahi Techno Glass, Japan) designed for low-energy photon beams.
A tin filter (Dose Ace, glass dosimeter; Asahi Techno Glass, Shizuoka, Japan) compensates for the high response of low-energy photons  2.F | Comparison of the radiation dose between the 10 cm ionization chamber, computed tomography dose index volume, and radio-photoluminescence glass dosimeters We compared the radiation dose between the 10 cm ionization chamber and computed tomography dose index volume (CTDIvol) of the console displayed dose using CT equipment because we had to check the accuracy of the CTDIvoi of the console displayed dose.
CTDIvol was good linearity of reference dose with the ionization chamber. We also compared the radiation dose between the 10 cm ionization chamber and radio-photoluminescence glass dosimeters using X-ray of general radiographic equipment due to the need to check the radio-photoluminescence glass dosimeters traceability. The radio-photoluminescence glass dosimeters were good for linearity of reference dose with the ionization chamber.

2.G | Size-specific dose estimate calculations
Size-Specific Dose Estimates (SSDE) were measured using the American Association of Physicists in Medicine (AAPM) Report 204-that is, to use of the anteroposterior (AP) parameter as a measurement of body thickness from anterior to posterior and the lateral (LAT) (c) parameter as a measurement of the body thickness from left to right and the summation of the AP and lateral dimensions and multiply the respective conversion factors with the CTDIvol at 16 cm.

2.H | Evaluation of organ and effective doses
Various tissues or organs were assigned an effective dose (ED) based on the definitions of the international commission on radiological protection (ICRP) by using three pediatric anthropomorphic phantoms. The ED was defined as follows: where W T was the tissue-weighing factor and W R was the radiation  Values of P < 0.05 were considered to be statistically significant.

2.J | Statistical analysis
Statistical analyses were with the free statistical software "R"

3.B | Radiation doses of whole-body exposure
In Tables 2-4, we present the radiation dose for each tissue or organ of the three phantoms; the noise index was 12 and TCM was applied

3.D | Image noise variations
Variations in the image noise on nonenhanced scans with TCM and a noise index of 12 are shown in Table 5

| DISCUSSION
We identified the radiation dose by scanning at 80, 100, and 120 kVp with TCM using anthropomorphic phantoms for a newborn, and a 1-year-old, and a 5-year-old child. Comparison of the radiation dose showed that at these tube voltages the delivered radiation dose was not significantly different.
Yu, et al. 9 reported that the noise level in an image obtained at 120 kVp is to be matched, and the potential for dose reduction at lower tube potentials is limited or nonexistent. For the 10 cm phantom, radiation dose is reduced by 12% at 80 kVp and by 8% at 100 kVp compared with the dose at 120 kVp. For the 25 cm phantom, a 29% dose increase is required at 80 kVp to match the noise level at 120 kVp. We used the three types of anthropomorphic T A B L E 2 Radiation dose calculation with SD-based TCM (SD 12) during chest and abdominal CT of a newborn anthropomorphic phantom. Life expectancy of children is longer than that of adults, their CT studies require the accurate evaluation of the whole-body exposure.
The dose parameters routinely displayed on scanner consoles include CTDI vol and DLP. As they are based on measurements in standard CT dose phantoms that are 16 or 32 cm in diameter. The effective dose is thought to correlate best with the overall stochastic radiation risk. 15 The dose calculations based on CTDI or DLP are readily avail- In the angular-modulation techniques they automatically adjust the tube current for each projection angle to attenuation of the patients to minimize x-rays in projection angle. In the z-axis-modulation technique, the system determines the tube current by using the patient localizer radiograph projection data and set of empirically determined noise projection coefficients by using reference technique. In earlier studies 22,23 ,3D TCM was on effective method for reducing the radiation dose delivered to patients and it could also reduce the radiation T A B L E 4 Radiation dose calculation with SD-based TCM (SD 12) during chest and abdominal CT of a 5-year-old anthropomorphic phantom.  When CT studies are performed in children, every effort must be made to select the optimal scanning protocols because they are more radiosensitive than adults and the radiation-induced stochastic effects are prolonged. Knowing and understanding the ALARA (as low as reasonably achievable) concept is imperative for making informed decisions regarding linical CT, research protocols, and longterm risk assessments.
The difference in the organ dose of low and standard tube voltage was minimal using pediatric phantoms. Shimonobo et al. 24 reported that among the pediatric phantoms there was no statistically significant difference in the mean surface and center dose at 80, 100, and 120 kVp until image noise level was maintained. In our results, surface dose of skin and center dose of lungs had no significant difference among the different tube voltages. Especially under the 5-year-old children it may not be on influence during different tube voltages.
Efforts of the AAPM to refine CTDIvol as SSDE with patient diameter are fairly accurate, with 10-20% variability. 25 In our study, the measured radiation dose was similar to the SSDE during lower tube voltage, it may be recommended to estimate SSDE on the basis of the conversion factors provided in AAPM Report 204.
Our study has some limitations. Firstly, we used anthropomorphic phantoms of newborn, 1-year-old and 5-year-old children to focus on routine and follow-up studies in pediatric patients. As they grow older, their body size increases and the results may be different. Secondly, our studies were performed on a single CT scanner model, from a single manufacturer. The relationship among the tube voltages, image noise, radiation dose, and phantom size may depend to some degree on the CT scanner specifications that may vary among manufacturers. Thirdly, we performed one helical scan for the new born, 1 year-old, and 5 year-old phantom at each tube voltage.
Lastly, we performed CT scan at one time for individual anthropomorphic phantoms.

| CONCLUSION S
Our anthropomorphic phantom study of the relationship between the radiation dose at different tube voltages in nonenhanced pediatric CT examinations suggests that the radiation dose at different low tube voltages is not significantly different when TCM is used.