Fetal radiation dose of four tube voltages in abdominal CT examinations during pregnancy: A phantom study

Abstract This study aimed to compare the dose and noise level of four tube voltages in abdominal computerized tomography (CT) examinations in different abdominal circumference sizes of pregnant women. Fetal radiation doses were measured with two anthropomorphic pregnant phantoms and real‐time dosimeters of photoluminescence sensors using four tube voltages for abdominal CT. The noise level was measured at the abdomen of two anthropomorphic pregnant phantoms. In the large pregnant phantom, the mean fetal doses performed using 120 and 135 kV were statistically significantly lower than the lower tube voltages (P < 0.05). In the small pregnant phantom, the mean fetal dose performed by 100, 120, and 135 kV was significantly lower than the lowest tube voltage tested (P < 0.05). The ratios of the peripheral mean dose to the centric mean dose showed that the ratios of 80 kV were the highest and those for 135 kV were the lowest in both pregnant phantoms. The ratios of the peripheral mean dose to the centric mean dose decreased as the tube voltage increased. Compared with low tube voltages, high tube voltages such as 120 and 135 kV could reduce radiation doses to the fetus without compromising the image uniformity in abdominal CT examinations during pregnancy. On low tube voltage protocols, the dose near the maternal skin surface may be increased in large pregnant women because of reduced penetration of the x rays.

tube voltages that were being used, ranging from low to high.
Although there are some reports that claim the use of a low tube voltage is helpful for radiation dose reduction in pediatric 12,13 and adult 14,15 CT, it is unclear whether the use of a low tube voltage is helpful for radiation dose reduction in fetuses. Thus, investigating the influence of the dose distribution inside the abdomen during pregnancy by varying tube voltages is important in the optimization process for the CT of pregnant women.
The abdominal circumference size of a pregnant woman changes considerably during pregnancy, which makes it important to estimate accurately the doses inside her abdomen. Many studies [16][17][18][19][20][21][22] have investigated the dose evaluation for fetuses. Although some previous studies have concentrated on determining fetal doses in radiological examinations using measurements in an anthropomorphic phantom or using Monte Carlo simulations, there are no reports about concrete dose reduction methods for fetuses. Thus, the focus of this study is to compare the fetal dose under the condition of the same noise as in the fetal region.
In this study, we compared the dose and noise level of four tube voltages in abdominal CT examinations in different abdominal circumference sizes of pregnant women.

2.A | CT equipment and scan parameters
We used a 320-multidetector CT (Aquilion ONE, Canon Medical Systems, Otawara, Japan) in this study. This study determined the scan parameters based on data collected from the Japanese nationwide dose survey 11 of CT for fetal skeletal dysplasia and was performed under the following tube voltages: 80, 100, 120, and 135 kV. Measurements of each tube voltage were obtained in the x-, y-, and z axes with tube current modulation (TCM) (Volume EC; Canon Medical Systems). To maintain a consistent image noise level for TCM [standard deviation (SD) setting], the CT equipment determined the tube current by the scout view data. The other scan parameters are presented in Table 1. The CT equipment generated images at the following reconstruction settings for noise measurements: SD setting of 31 at a thickness of 0.5 mm; transverse slices, 5 mm; current range, 10-900 mA; and reconstruction filter, FC12. The images were reconstructed using an iterative technique, namely, the adaptive iterative dose reduction three-dimensional (AIDR3D) technique, and the AIDR3D strength was weak.

2.B | Pregnant model phantom and dosimeter placement
A pregnant model phantom was constructed using an anthropomorphic phantom (Alderson Rando phantom) and two differently sized custom-made abdomen phantoms simulating pregnancy (Kyoto Kagaku Co. Ltd., Kyoto, Japan) ( Fig. 1), hereafter referred to as the large and small pregnant phantoms. The custom-made abdomen phantoms during pregnancy were constructed with a polyurethane resin. The specific gravity of the polyurethane resin was 1.06. The size and shape of the polyurethane resin was designed based on the abdominal size and shape collected from CT examinations of 18 pregnant patients (gestational ages of 8 to 39 weeks) in one hospital.
The abdominal circumferences at the umbilicus of the large and small pregnant phantoms were 80 and 95 cm, respectively. The fetal radiation doses were measured using real-time dosimeters (RTDs) (RD-1000, TORECK Corporation, Kanagawa, Japan) (Fig. 2), which is a very simple and immediate reading dose measurement method without heating and annealing, compared to the conventional methods, such as thermoluminescent dosimeters 16 and radiophotoluminescence glass dosimeters. 23 This RTD comprised photoluminescence sensors (Y 2 O 2 S: Eu, Sm), an optical fiber cable, a photodiode, and a digital display that included the power supply, and can measure using a maximum of four sensors in one measurement. [24][25][26][27] The photoluminescence sensor size had a cylindrical shape with dimensions φ4.1 × 11.5 mm. RTDs were implanted at 11 and six points at the umbilical level of the large and small pregnant phantoms, respectively (Fig. 3). For the 11 measurement points on the large pregnant phantom, one point was measured as the central dose and 10 points were measured as the peripheral dose. Analogously, for the six measurement points on the small pregnant phantom, one point was measured as the central dose and five points were measured as the peripheral dose. Each measurement was performed three times to reduce random error. The scanning start angle of the x-ray tube for each measurement was guaranteed the same by using the orbital synchronized scanning technique.

2.D | Measurement of noise levels
The image noise was measured by drawing uniform regions of interest (ROIs) in areas of the large and small pregnant phantoms in a soft tissue algorithm and expressed as the SD of Hounsfield units (HU). For each tube voltage, three 30-mm 2 circular ROIs were drawn at three separate slice levels and placed manually at the same visible location on the image (Fig. 4). The mean noise was calculated at each slice level. The RTD coverage spanned one image (thickness of 5 mm). "Three slices" for the noise measurement corresponded to the dosimeter's longitudinal coverage.

2.E | Statistics
The statistical significance of dose and image noise differences among the four tube voltages was evaluated using the Steel-Dwass procedure, which was performed using the statistical software MEPHAS (Osaka University, Osaka, Japan). A P < 0.05 was considered to be statistically significant.

3.A | Radiation dose
The mean fetal radiation doses of 11 points measured using the four tube voltages for abdominal CT examinations for the large and small pregnant phantoms are shown in Table 2. In the large pregnant phantom, the mean fetal doses measured at 120 and 135 kV were statistically significantly lower than the tube voltages of 80 and 100 kV (P < 0.05). In the small pregnant phantom, the mean fetal doses measured at 100, 120, and 135 kV were significantly lower than the tube voltage of 80 kV (P < 0.05). For the mean fetal radiation doses measured using the four tube voltages, the mean fetal dose measured at 120 kV was the lowest for both pregnant phantoms. The ratios of the peripheral mean dose to the centric mean dose measured using the four tube voltages are shown in Table 3.

3.B | Image noise
The image noise at the three points measured using the four tube voltages for the abdominal CT examinations for the large and small pregnant phantoms are shown in Table 4. There were no statistically significant differences among the four tube voltages in the SD for all ROIs in the images using the Steel-Dwass procedure.

| DISCUSSION
This study showed that the use of a high tube voltage protocol enables substantial fetal radiation dose reduction for abdominal CT  reported that for patients with a body width >30 cm, dose optimized protocols required 120 kV combined with a more aggressive reduction in tube current. The abdominal circumferences of pregnant women are typically larger than those of children and general adults.
In this study, the mean fetal dose using high voltages of 120 and 135 kV for the large pregnant phantom were significantly lower than those using 80 and 100 kV. On the other hand, the mean fetal dose for the small pregnant phantom was significantly lower when using ≥ 100 kV. If the abdominal circumference of a pregnant women is over 95 cm and the setting is <120 kV or the abdominal circumference is less than 80 cm and the setting is <100 kV, an increase in fetal radiation dose may result.
The increasing fetal radiation dose by low tube voltage protocols is caused by the peripheral dose near the maternal skin surface being higher than when compared with settings ≥ 120 kV for the large pregnant phantom and with settings ≥ 100 kV for the small pregnant phantom. Moreover, the ratios of the peripheral mean dose to the centric mean dose inside phantoms were high at lower tube voltages. The cause is thought to be decreased penetration in the pregnant woman as a result of x-ray absorption near the skin surface when following low tube voltage protocols. A previous study 30 reported that there was no correlation between fetal depth (distance from skin to fetus) and abdominal circumferences or gestational ages. There is a possibility that the position of the fetus is close to the maternal skin surface regardless of the abdominal circumferences and gestational ages. Therefore, we think that most noncontrast F I G . 5. Absorbed radiation dose distributions using the four tube voltages for the large pregnant phantom.
F I G . 6. Absorbed radiation dose distributions using the four tube voltages for the small pregnant phantom.
T A B L E 4 Image noise using the four tube voltages for abdominal CT examinations during pregnancy for large and small pregnant phantoms. abdominal CT examinations during pregnancy should not use the low tube voltage protocols, but should instead use 120 or 135 kV.
A major advantage of a low tube voltage protocol is the improved image contrast by the attenuation of iodinated structures which steeply increases as the effective energy of the x-ray spectrum approaches the iodine k-edge (33.2 keV) with the use of a contrast medium. 12,13  regard to the abdomen/pelvis studies for average patient sizes. It is unclear whether the same results will be obtained for the abdomen/ pelvis studies for above-or below-average patient sizes.