General information: This was a prospective clinical study and was approved by the Ethics Committees of our institution. The informed consent was signed by children’s parents. From Feb 26th, 2018 to Mar 30th, 2018, the patients undergoing abdominal CTA were continuously collected into the study group. Exclusion criteria included: the body weight greater than 28.0 kg. Children with matching age and weight who underwent abdominal CTA using the conventional CT scan protocol before June 1st, 2015 were selected from hospital database into the control group for comparison purpose.
Instruments and equipment: For the study group, a 256-row CT scanner (Revolution CT, GE Healthcare, USA) was used with the following scan parameters: tube voltage of 70kVp, helical pitch of 1.375:1 and tube rotation speed of 0.35 s, A set of fixed tube currents were used and were adjusted to obtain the following patient weight-dependent radiation dose (calculated as CTDIvol) [15]: 0.92 mGy for 0–12 kg with 260 mA; 1.22 mGy for 12.1–20 kg with 345 mA; and 1.52 mGy for 20.1–28 kg with 430 mA. The images were reconstructed at 0.625 mm slice thickness using the second-generation adaptive statistical iterative reconstruction (ASIR-V) at 50% strength (50%ASIR-V) and a standard reconstruction kernel. For the control group, a 64-row CT scanner (Discovery CT750HD, GE Healthcare, USA) was used with the conventional scan protocol: tube voltage of 100 kV, helical pitch of 1.375:1 and tube rotation speed of 0.4 s. The tube current was set by using the automatic tube current modulation (ATCM) in the range of 10–700 mA during the scan to obtain age-based image noise index (NI) settings: NI = 11HU for children with age of 0–12 months; NI = 13HU for 1–2 years old, and NI = 15HU for 3–14 years old. The images in the control group were reconstructed at 0.625 mm slice thickness using ASIR algorithm with 50% strength (50%ASIR) and a standard reconstruction kernel. For those children who were too young to cooperate, sedation with oral chloral hydrate (10%, 0.5 ml/Kg) was applied before the scanning in both groups.
Enhanced CT protocol: A peripheral venous cannula was pre-placed in the superficial vein of dorsum of the hand. A 22G needle was used for children with body weight > 15.1 kg and a 24G needle was used for children with weight < 15.0 kg. An iodinated contrast agent (270 mg I/ml iodixanol; GE healthcare, American) was administered in both groups using a single-head power injector. The contrast medium dosage in the control group was calculated according to the body weight of each child: 1.8 ml/kg for 3–5 kg, 1.6 ml/kg for 5–10 kg, 1.4 ml/kg for 10–15 kg and 1.2 ml/kg for 15–28 kg, while the contrast medium dosage in the study group was reduced by 40%: 1.1 ml/kg for 3–5 kg, 1.0 ml/kg for 5–10 kg, 0.8 ml/kg for 10–15 kg and 0.7 ml/kg for 15–28 kg. Flow rate was adjusted according to a fixed injection time of 15 s and contrast enhanced scan started at 17 s after the start of contrast injection.
All images were transmitted to a GE AW4.7 workstation (GE Healthcare, WI, USA) for data measurement and image analysis. All children related information and scanning parameters were hidden during the image analysis process, and 3D post-processing images, such as multi-planner reformation (MPR), and Volume Rendering (VR), were generated and used to display the vessels. Two pediatricians (with 14 years and 7 years of CT diagnostic experience) evaluated the image quality according to the scoring standard together with consensus, and they could adjust the image display window width and window level to the level deemed appropriate. Image evaluation included the subjective scoring for vessel display and objective measurement of CT attenuation value and noise value of image.
Subjective image quality evaluation: The subjective image quality assessment focused mainly on the sharpness and clearness of the edges of large vessels and the ability to display branches of small mesenteric arteries. The subjective quality evaluation of the large vessels including the descending aorta and its first and second branches was performed on a 5-point scale as follows (with 3–5 stands for a satisfying imaging quality): 5 point, the edge of the vessel is clearly displayed, with a very good contrast; 4 point, the edge is clear, with a good contrast; 3 point, the blood vessel is clear, but not smooth enough, the diameter of the lumen can be accurately measured; 2 point: the diameter of the lumen cannot be accurately measured, but the shape of the blood vessel can be judged; 1 point: the shape of the blood vessel cannot be judged. The ability for vascular display was evaluated by the observed number of branches of small mesenteric arteries. The number of small mesenteric artery branches were observed on the VR images (more than 50% of the branch to be displayed), and the smallest branch (with at least 2 branches displayed) could be observed. The image quality was jointly evaluated by two readers. In case of any inconsistency, the final score was given after consultation.
Objective measurement: the CT value and standard deviation (SD) of the interested area of descending aorta (Ao) and back muscle (Mu) of hepatic hilar section were measured by the two observers together after the subjective evaluation, SD values were used to represent image noise. The signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of the descending aorta were calculated using the following formula: SNR = CT density(Ao) / SD(Ao), CNR = (CT density(Ao)-CT density(Mu))/ ((SD(Ao)་SD(Mu))/2).
Statistical analysis: the general information for the patients including height, weight, gender and age were recorded in detail. The scanning and contrast agent parameters including radiation dose (volumetric CT dose index (CTDIvol), dose-length-product (DLP)), contrast medium volume, iodine dose, contrast injection rate and IV pressure were recorded. The objective noise and subjective score data including CT and SD values of Ao and Mu, SNR, CNR, and subjective scores were represented as mean ± standard deviation. Paired t-test was performed to assess for the continuous data that followed normal distribution to evaluate whether there were differences between the two groups, Mann-Whitney U test was performed for the discrete data such as the subjective image quality scores and for the continuous data that did not follow the accord with normal distribution. The spearman correlation coefficients between flow rate and injection pressure in two group were evaluated. All the statistical analyses were performed using SPSS17.0 software, P < 0.05 was considered as statistically significant.