Abstract
Objectives:
To assess the feasibility of whole-body dual-energy computed tomographic angiography (DECTA) at 40 keV with 50% reduced iodine dose protocol.
Methods:
Whole-body CTA was performed in 65 patients; 31 of these patients underwent 120 kVp single-energy computed tomographic angiography (SECTA) with standard iodine dose (600 mgI/kg) and 34 with 40 keV DECTA with 50% reduced iodine dose (300 mgI/kg). SECTA data were reconstructed with adaptive statistical iterative reconstruction of 40% (SECTA group), and DECTA data were reconstructed with adaptive statistical iterative reconstruction of 40% (DECTA-40% group) and 80% (DECTA-80% group). CT numbers of the thoracic and abdominal aorta, iliac artery, background noise, signal-to-noise ratio (SNR), and arterial depiction were compared among the three groups. The CT dose index volumes (CTDIvol) for the thorax, abdomen, and pelvis were compared between SECTA and DECTA protocols.
Results:
The vascular CT numbers and background noise were found to be significantly higher in DECTA groups than in the SECTA group (p < 0.001). SNR was significantly higher in the order corresponding to DECTA-80%, SECTA, and DECTA-40% (p < 0.001). The arterial depiction was comparable in almost all arteries; however, intrapelvic arterial depiction was significantly worse in DECTA groups than in the SECTA group (p < 0.0001–0.017). Unlike the pelvic region (p = 0.055), CTDIvol for the thorax (p < 0.0001) and abdomen (p = 0.0031) were significantly higher in the DECTA protocol than in the SECTA protocol.
Conclusion:
DECTA at 40 keV with 50% reduced iodine dose provided higher vascular CT numbers and SNR than SECTA, and almost comparable arterial depiction, but had a degraded intrapelvic arterial depiction and required a larger radiation dose.
Advances in knowledge:
DECTA enables 50% reduction of iodine dose while maintaining image quality, arterial depiction in almost all arteries, vascular CT numbers, and SNR; however, it does not allow clear visualization of intrapelvic arteries, requiring a slightly larger radiation dose compared with SECTA with standard iodine dose.
Introduction
Computed tomographic angiography (CTA) plays an important role in understanding vascular anatomic features, diagnosis of aortic diseases,1 therapy planning for endovascular aortic aneurysm repair (EVAR) and transcatheter aortic valve implantation,2,3 and surveillance after EVAR.4,5 However, patients with aortic diseases who need CTA often have pre-existing renal insufficiency, which increases their risk for contrast-induced nephropathy (CIN).6 For example, especially in patients who have had EVAR, multiple repeated CTA studies for treatment follow-up are recommended to detect possible complications, including endoleak.7 The risk of CIN has been related to the amount of contrast material; therefore, a CTA protocol requiring the smallest amount of contrast material is needed in clinical practice.
When applying low energy levels, the virtual monochromatic images (VMIs) generated from a dual-energy CT (DECT) data set have the potential to provide substantially higher CT attenuation. In particular, Shuman et al8 reported that VMIs at 50 kilo-electron volt (keV) with iterative reconstruction resulted in similar aortic CT attenuation, signal-to-noise ratio (SNR), and contrast-to-noise ratio with 70% reduced iodine dose in abdominal dual-energy CTA (DECTA) compared with the conventional single-energy CTA (SECTA) at 120 kilovolt peak (kVp) with standard iodine dose.
On the other hand, clinical utility of VMIs is often limited by the expanse of a significant increase of the image noise at lower energy levels.9,10 Moreover, the image noise tends to be higher in the pelvic region compared with the thorax and abdomen.11 We wondered about the influence of anatomic locations on image quality in DECTA at lower energy levels; however, to our knowledge, no study has yet evaluated the CTA image quality in this regard. Thus, the purpose of this study was to compare the quantitative and qualitative image qualities of whole-body CTA between SECTA at 120 kVp with standard iodine dose and DECTA at 40 keV with 50% reduced iodine dose by anatomical locations.
Methods and materials
Patients
Our institutional review board has provided approval for this prospective study. Written informed consent was obtained from all patients. Seventy-two consecutive patients who underwent whole-body CTA for evaluation of aortic diseases between August 2018 and July 2019 were included in this study. Out of these 72 patients, 7 were excluded because of different rotation times (n = 3), inadequate scan timing (n = 2), no raw data (n = 1), and different slice thickness (n = 1). The final cohort consisted of 65 patients (55 males and 10 females; mean age, 72.4 ± 8.4 years; range, 52–88 years; mean body weight, 63.7 ± 10.4 kg; range, 39.0–95.0 kg; mean body mass index, 24.2 ± 3.3 kg/m2; range, 16.8–36.2 kg/m2) who were included for data analysis. Most of the clinical diagnoses were abdominal aortic aneurysm after EVAR (n = 31), abdominal aortic aneurysm after artificial blood vessel replacement surgery (n = 9), thoracic aortic aneurysm after thoracic endovascular aortic aneurysm repair (TEVAR) (n = 8), abdominal aortic aneurysm (n = 8), aortic dissection (n = 5), thoracic and abdominal aneurysms after TEVAR and EVAR (n = 2), thoracic aortic aneurysm after artificial blood vessel replacement surgery (n = 1), and superior mesenteric arterial dissection (n = 1).
Phantom study
Before clinical scanning, we performed a phantom study to determine the threshold in bolus-tracking technique. The self-produced phantom consisted of water in a hermetic container and six syringes with different densities of diluted contrast material. The phantom was scanned by a fast kilovoltage-switching DECT scanner (Discovery CT750 HD; GE Healthcare, Milwaukee, WI) at 140 kVp, which corresponds to the tube voltage in real-time fluoroscopic monitoring scans. The scan parameters were as follows: tube current, 10 mA in SE and 365 mA in DE scans; rotation time, 0.5 rot/s; slice thickness, 5.0 mm with fluoroscopic scan. The DE scanning data set was reconstructed at 40 keV, which was used in this study. A circular region of interest (ROI) was placed on an internal syringe at 140 kVp and VMI at 40 keV, respectively, and CT numbers were measured. From these data, we obtained the following approximation formula, y = 0.0026x2 + 3.071x + 6.9292, where x means CT numbers at 140 kVp and y means CT numbers of VMIs at 40 keV. We calculated the proposed threshold for VMIs at 40 keV, which corresponded to 100 Hounsfield units (HUs) at 120 kVp scanning, and the threshold was 29.6 HU. Therefore, we used 30 HU as a threshold in DECTA scanning in this study.
Dual-energy CT technique and contrast injection
All examinations were performed with a fast kilovoltage-switching DECT scanner (Discovery CT750 HD; GE Healthcare, Milwaukee, WI). The patients were randomized to undergo scanning with SECT or DECT: SECTA, CTA scanned by SECT at 120 kVp with 600 mgI/kg contrast material (300 mg iopamidol per milliliter [mL]), and DECTA, CTA scanned by DECT at 40 keV with 300 mgI/kg contrast material (300 mg iopamidol per mL). The CT imaging parameters in SECTA were as follows: tube voltage, 120 kVp; noise index, 10.0 HU at 5 mm slice thickness; tube current, variable; detector configuration, 64 detectors with 0.625 mm section thickness; beam collimation, 40 mm; rotation time, 0.5 s; pitch, 0.984:1; scan field of view (FOV), large body; and display FOV, 36 cm. Meanwhile, CT imaging parameters in DECTA were as follows: noise index, 12.0 HU at 5 mm slice collimation; tube current, variable (GSI Assist; GE Healthcare, Milwaukee, WI); detector configuration, 64 detectors with 0.625 mm section thickness; beam collimation, 40 mm; rotation time, 0.5 s; pitch, 0.984:1; scan FOV, large body; and display FOV, 36 cm.
The contrast material was intravenously injected at 4 ml/s using a commercially available power injector. A circle with a diameter of 15–20 mm was placed as an ROI in the descending aorta at the level of the bronchial carina. Real-time fluoroscopic monitoring scans (140 kVp, 10 mA) were initiated 5 s after contrast injection. The contrast injection was discontinued when the bolus-tracking program (SmartPrep; GE Healthcare) detected contrast enhancement reaching 30 HU; this was followed by 20 ml saline injection at 4 ml/s. Diagnostic CT scanning was initiated with an additional delay of 5 s after a bolus-tracking program detected a threshold attenuation. CT scans in the craniocaudal direction were obtained from the supraclavicular fossa to the pubic symphysis.
Image reconstruction
Raw projection of CTA data at 1.25 mm section thickness with 50% overlap was reconstructed with an adaptive statistical iterative reconstruction at 40% in SECTA (SECTA group) and DECTA (DECTA-40% group) and 80% in DECTA (DECTA-80% group). The reconstructed axial CTA images were reformatted to three-dimensional volume-rendered (VR) and multiplanar reformatted (MPR) CTA images at a workstation (Advantage Windows 4.6; GE Healthcare). VR reconstruction parameters were as follows: a linear threshold with a lower threshold value of 125 HU set at 0% opacity (completely transparent) and an upper threshold value of 600 HU set at 100% opacity (completely opaque). The z-axis was selected for rotating three-dimensional observation at every 10° over 360° of rotation, which yielded 36 VR images. MPR images were reconstructed in the coronal plane with a 2.5 mm section thickness and no intersectional gap.
Quantitative image analysis
An experienced radiologist (F.N., with 4 years of post-training experience at interpreting body CT images) measured the CT number of the thoracic aorta (mean CT values at ascending aorta, aortic arch, and descending aorta), abdominal aorta (mean CT values at the upper, middle, and lower abdominal aorta), and iliac arteries (mean CT values at bilateral common iliac arteries) using commercially available digital imaging software and a communications in medicine (DICOM) viewer. The radiologist placed a circular ROI of 5–30 mm in diameter on axial images, encompassing as much of the vascular lumen as possible while avoiding the vascular walls, calcification, thrombus, medical devices, and artifacts. For each patient, one standard deviation (SD) of the CT numbers of the homogeneous subcutaneous fat tissue at the level of the carina, upper pole of the left kidney, and cranial aspect of the femoral head representative of thoracic, abdominal, and pelvic regions was determined as the background noise. An SNR was calculated by dividing the CT number of each vessel by background noise at corresponding anatomic regions.
Qualitative image analysis
The experienced radiologists (Y.N. and N.K., with 9 and 8 years of post-training experience at interpreting body CT images, respectively), who was blinded of the CTA protocols, independently reviewed the CTA images and then in consensus. The radiologists randomly examined the axial, MPR, and VR images. Preset window settings of axial and MPR images were initially fixed with an optimal window level of 140 HU and window width of 680 HU.12
The radiologists graded the arterial depiction of the brachiocephalic, common carotid, subclavian, bronchial, internal thoracic, intercostal, common hepatic, proper hepatic, splenic, left gastric, gastroduodenal, inferior phrenic, superior mesenteric, inferior mesenteric, renal, lumbar, common iliac, external iliac, internal iliac, iliolumbar, superior gluteal, inferior gluteal, obturator, and inferior epigastric arteries using a 5-point rating scale. A grade of 5 meant all vascular segments from the trunk to the subsegmental peripheral artery were clearly visualized; a grade of 4 meant intermediate between Grades 5 and 3; a grade of 3 meant that nearly half of all vascular segments were clearly visualized; a grade of 2 meant intermediate between Grades 3 and 1; and a grade of 1 meant that none of the vascular segments were clearly visualized. The radiologists did not evaluate the arterial depiction of bronchial and intercostal arteries in post-surgical patients for thoracic aortic aneurysm, that of inferior phrenic and lumbar arteries in post-surgical patients for abdominal aortic aneurysm, or that of internal iliac, iliolumbar, superior gluteal, inferior gluteal, and obturator arteries in patients with internal iliac artery after coil embolization. The radiologists further evaluated the image quality for sharpness, granulation, noise, and diagnostic acceptability using a 5-point scale: a grade of 5 meant “excellent”; a grade of 4 meant “good”; a grade of 3 meant “acceptable”; a grade of 2 meant “suboptimal”; and a grade of 1 meant “unacceptable.”
Radiation exposure
The CT dose index volume (CTDIvol) and dose–length product (DLP) from the dose report were recorded. On the SE scans, tube current modulation adjusted the tube current for the size and shape of individual patients, who were monitored on a scout scan accounting for all three dimensions. Conversely, the CTDIvol was fixed at 10.68, 15.02, or 23.56 mGy on DE scans depending on the patient’s body size using GSI Assist. DLP and CTDIvol for the thorax, abdomen, and pelvis were estimated with the following formula:
where DLPT, A, or P, CTDIvolT, A, or P, and scan rangeT, A, or P are the DLP, CTDIvol, and scan range (cm) for the thorax (T), abdomen (A), or pelvis (P); Ck, the tube current recorded in DICOM tags of CT images; l, the most caudal image number for the thorax; m, the most caudal image number for the abdomen; and n, the most caudal image number of the pelvis. The thorax represents the coverage from the most cranial image of each image series to the image for the lower aspect of T12. The abdomen represents the coverage from the upper aspect of L1 to the lower aspect of L5. The pelvis represents the coverage from the upper aspect of the sacrum to the most caudal image of the series.11
Statistical analysis
Statistical analyses were performed using the MedCalc statistical software for Windows (MedCalc software v. 19.2; Mariakerke, Belgium). Welch’s and Fisher’s tests were performed in order to evaluate the differences between the SECTA and the DECTA groups in terms of the patients’ age, sex, height, body weight, body mass index, volume of contrast material injected, and weight of iodine injected. One-way analysis of variance and a post-hoc Tukey test were conducted to evaluate differences in terms of the CT numbers of arteries, background noise, and SNR among the three groups. The Welch’s test was conducted to evaluate differences between the SECTA and the DECTA groups in regard to the CTDIvol and DLP for the thorax, abdomen, and pelvis. The Kruskal–Wallis test was utilized to compare the qualitative grades among the three groups. When a significant difference was found in the three groups, pairwise comparisons were performed with the Mann–Whitney U test, and a stricter p < 0.017, introducing the Bonferroni correction, was considered significant. A p-value of less than 0.05 was considered to be significant.
Results
Patient demographics
Patient demographics between SECTA and DECTA groups are summarized in Table 1. The mean height of the patients was found to be significantly higher in the DECTA group than in the SECTA group (p = 0.02). Meanwhile, the mean contrast volume and iodine weight were significantly smaller in the DECTA group compared with the SECTA group (p < 0.0001). Between SECTA and DECTA groups, no significant difference was found in patients’ age (p = 0.43), sex (p = 0.17), body weight (p = 0.21), or body mass index (p = 0.98).
Table 1.
Patient demographics between single-energy and dual-energy CT groups
| Parameter | SECT (120 kVp with 600 mgI/kg) | DECT (40 keV with 300 mgI/kg) | p-value |
|---|---|---|---|
| Patient no. | 31 | 34 | N.A. |
| Age (y) | 71.6 ± 8.3 (52.0–86.0) | 73.2 ± 8.5 (56.0–83.0) | 0.43 |
| Male:Female | 24:7 | 31:3 | 0.17 |
| Height (cm) | 159.9 ± 6.4 (147.0–173.0) | 164.3 ± 8.4 (139.0–176.0) | 0.02a |
| Body weight (kg) | 62.0 ± 9.6 (39.0–78.0) | 65.3 ± 11.1 (36.0–78.0) | 0.21 |
| Body mass index | 24.2 ± 3.0 (16.8–29.5) | 24.2 ± 3.5 (18.5–36.2) | 0.98 |
| Contrast volume (ml) | 93.7 ± 22.1 (61.0–144.0) | 42.1 ± 8.2 (23.0–59.0) | <0.0001a |
| Iodine weight (g) | 28.1 ± 6.6 (18.3–43.2) | 12.6 ± 2.5 (6.9–17.7) | <0.0001a |
DECT, Dual-energy computed tomography; N.A, Not applicable; SECT, Single-energy computed tomography.
Data are means ± 1 standard deviation with ranges in parentheses.
p < 0.05, significant difference.
Quantitative image analysis
The CT numbers, background noise, and SNRs in the three groups are summarized in Table 2. The mean CT numbers of the thoracic aorta, abdominal aorta, and iliac arteries were found to be significantly greater in the DECTA-40% and DECTA-80% groups than in the SECTA group (p < 0.001). The background noises of the thoracic, abdominal, and pelvic regions were also found significantly greater in the DECTA-40% group than in the SECTA and DECTA-80% groups, and those of the abdominal and pelvic regions were significantly greater in the DECTA-80% group than in the SECTA group (p < 0.001). The mean SNRs of the thoracic aorta, abdominal aorta, and iliac arteries were recorded to be significantly greater in the DECTA-80% group than in the SECTA and DECTA-40% groups, and those of the abdominal aorta and iliac arteries were also significantly greater in the SECTA than in the DECTA-40% group (p < 0.001).
Table 2.
CT number, background noise, and signal-to-noise ratio in the three groups
| Anatomic site | SECTA | DECTA-40% | DECTA-80% | p-value |
|---|---|---|---|---|
| Thoracic aorta | ||||
| CT number | 356.3 ± 47.3a (273.9–475.2) | 600.1 ± 89.2 (383.5–748.3) | 598.9 ± 89.4 (381.5–747.4) | <0.001 |
| Background noise | 11.5 ± 1.8 (7.7–16.2) | 21.4 ± 3.6b (15.6–28.1) | 12.6 ± 3.1 (7.3–22.7) | <0.001 |
| SNR | 31.7 ± 6.5 (16.9–48.5) | 28.7 ± 6.1 (19.9–43.2) | 50.4 ± 14.3c (23.8–82.0) | <0.001 |
| Abdominal aorta | ||||
| CT number | 345.9 ± 51.8a (199.4–446.8) | 581.4 ± 82.4 (379.8–734.2) | 580.1 ± 82.5 (376.6–733.9) | <0.001 |
| Background noise | 10.8 ± 1.9a (7.7–15.0) | 23.3 ± 5.3b (13.5–38.2) | 13.1 ± 3.6 (7.1–21.8) | <0.001 |
| SNR | 33.0 ± 6.9c (13.3–46.7) | 26.0 ± 6.3d (15.5–44.4) | 47.1 ± 13.0c (27.2–79.4) | <0.001 |
| Iliac arteries | ||||
| CT number | 343.6 ± 45.6a (245.1–456.3) | 546.5 ± 81.8 (370.8–708.2) | 543.5 ± 82.8 (363.5–704.9) | <0.001 |
| Background noise | 10.3 ± 1.5a (7.1–14.4) | 24.5 ± 5.6b (16.7–47.7) | 13.6 ± 3.1 (8.8–25.4) | <0.001 |
| SNR | 34.1 ± 7.1 (20.7–50.1) | 23.2 ± 5.8d (11.8–40.8) | 42.0 ± 12.0c (21.8–73.2) | <0.001 |
DECTA, Dual-energy computed tomographic angiography; SECTA, Single-energy computed tomographic angiography; SNR, Signal-to-noise ratio.
Data are means ± 1 standard deviation with ranges in parentheses.
Value was significantly lower than those with DECTA-40% and DECTA-80% groups (p < 0.05).
Value was significantly greater than those with SECTA and DECTA-80% groups (p < 0.05).
Value was significantly greater than those with SECTA and DECTA-40% groups (p < 0.05).
Value was significantly lower than those with SECTA and DECTA-80% groups (p < 0.05).
Qualitative image analysis
Arterial depiction
The arterial depictions on axial, MPR, and VR images are summarized in Tables 3–5. For axial images, the arterial depictions of the bronchial, intercostal, inferior phrenic, obturator, and inferior epigastric arteries were significantly worse in the SECTA group than in the DECTA groups (p < 0.001–0.0050), but the arterial depictions of the lumbar and iliolumbar arteries were significantly better in the SECTA group than in the DECTA groups (p < 0.001). No significant difference was detected in the other arteries (p = 0.24–1.00). For MPR images, the arterial depiction of the intercostal artery was significantly worse in the SECTA group than in the DECTA groups (p < 0.001), and that of the inferior phrenic artery was significantly worse in the SECTA group than in the DECTA-80% group (p = 0.028). On the other hand, the arterial depictions of the lumbar, iliolumbar, inferior gluteal, and inferior epigastric arteries were significantly better in the SECTA group than in the DECTA groups (p < 0.001–0.0045). No significant difference was detected in the other arteries (p = 0.11–1.00). For VR images, the arterial depictions of the inferior phrenic, lumbar, iliolumbar, and superior gluteal arteries were found to be significantly worse in the DECTA groups than in the SECTA group (p < 0.001–0.0066); the same goes for the inferior gluteal artery, which was significantly worse in the DECTA-40% group than in the SECTA group (p = 0.017). For the left gastric artery, a significant difference was found using the Kruskal–Wallis test (p = 0.035); however, no significant difference was detected when pairwise Mann–Whitney U test was performed (p > 0.017). No significant difference was detected in the other arteries (p = 0.10–1.00) (Figures 1–3).
Table 3.
Arterial depiction on axial images
| Artery | SECTA | DECTA-40% | DECTA-80% | p-value |
|---|---|---|---|---|
| Brachiocephalic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Common carotid | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Subclavian | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Bronchial | 2.5 ± 1.4a | 3.6 ± 1.6 | 3.6 ± 1.6 | 0.0046 |
| Internal thoracic | 4.5 ± 0.6 | 4.5 ± 0.6 | 4.5 ± 0.6 | 0.90 |
| Intercostal | 3.2 ± 0.7a | 3.8 ± 0.9 | 4.0 ± 0.8 | <0.001 |
| Common hepatic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Proper hepatic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Splenic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Left gastric | 4.7 ± 0.8 | 4.7 ± 1.0 | 4.7 ± 1.0 | 0.86 |
| Gastroduodenal | 4.9 ± 0.4 | 4.8 ± 0.9 | 4.8 ± 0.8 | 0.60 |
| Inferior phrenic | 3.8 ± 0.7a | 4.4 ± 0.8 | 4.4 ± 0.9 | 0.0019 |
| Superior mesenteric | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Inferior mesenteric | 2.5 ± 1.8 | 1.8 ± 1.6 | 1.9 ± 1.6 | 0.24 |
| Renal | 4.9 ± 0.3 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Lumbar | 3.5 ± 0.8b | 2.8 ± 0.6 | 2.7 ± 0.6 | <0.001 |
| Common iliac | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| External iliac | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Internal iliac | 4.9 ± 0.1 | 4.6 ± 1.1 | 4.6 ± 1.1 | 0.57 |
| Iliolumbar | 4.9 ± 0.2b | 3.9 ± 1.2 | 3.9 ± 1.2 | <0.001 |
| Superior gluteal | 4.8 ± 0.4 | 4.5 ± 1.1 | 4.4 ± 1.1 | 0.36 |
| Inferior gluteal | 4.3 ± 0.8 | 4.4 ± 0.9 | 4.4 ± 0.9 | 0.60 |
| Obturator | 3.7 ± 0.8a | 4.5 ± 1.0 | 4.5 ± 1.0 | <0.001 |
| Inferior epigastric | 4.0 ± 0.7a | 4.5 ± 0.7 | 4.5 ± 0.6 | 0.0050 |
DECTA, Dual-energy computed tomographic angiography; SECTA, Single-energy computed tomographic angiography.
Data are means ± 1 standard deviation.
Value was significantly lower (p < 0.017) than those with DECTA-40% and DECTA-80% groups.
Value was significantly greater (p < 0.017) than those with DECTA-40% and DECT-80% groups.
Table 4.
Arterial depiction on multiplanar reformatted images
| Artery | SECTA | DECTA-40% | DECTA-80% | p-value |
|---|---|---|---|---|
| Brachiocephalic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Common carotid | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Subclavian | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Bronchial | 4.1 ± 1.1 | 3.6 ± 1.6 | 3.6 ± 1.6 | 0.50 |
| Internal thoracic | 4.5 ± 0.6 | 4.5 ± 0.6 | 4.5 ± 0.6 | 0.95 |
| Intercostal | 3.2 ± 0.7a | 4.0 ± 0.9 | 4.1 ± 0.8 | <0.001 |
| Common hepatic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Proper hepatic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Splenic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Left gastric | 4.9 ± 0.7 | 4.7 ± 1.0 | 4.7 ± 1.0 | 0.41 |
| Gastroduodenal | 4.9 ± 0.4 | 4.8 ± 0.4 | 4.8 ± 0.4 | 0.67 |
| Inferior phrenic | 3.8 ± 0.7b | 4.1 ± 0.9 | 4.3 ± 0.9 | 0.028 |
| Superior mesenteric | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Inferior mesenteric | 2.5 ± 1.8 | 1.8 ± 1.6 | 1.9 ± 1.7 | 0.24 |
| Renal | 4.9 ± 0.3 | 5.0 ± 0.0 | 5.0 ± 0.0 | 0.11 |
| Lumbar | 3.5 ± 0.8c | 2.8 ± 0.6 | 2.7 ± 0.6 | <0.001 |
| Common iliac | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| External iliac | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Internal iliac | 4.9 ± 0.1 | 4.6 ± 1.1 | 4.6 ± 1.1 | 0.23 |
| Iliolumbar | 4.9 ± 0.2c | 3.8 ± 1.2 | 3.8 ± 1.2 | <0.001 |
| Superior gluteal | 4.8 ± 0.3 | 4.5 ± 1.1 | 4.4 ± 1.1 | 0.45 |
| Inferior gluteal | 4.9 ± 0.2c | 4.5 ± 0.9 | 4.5 ± 0.9 | 0.0015 |
| Obturator | 4.7 ± 0.5 | 4.2 ± 1.1 | 4.3 ± 1.1 | 0.21 |
| Inferior epigastric | 4.9 ± 0.3c | 4.5 ± 0.6 | 4.5 ± 0.6 | 0.0045 |
DECTA, Dual-energy computed tomographic angiography; SECTA, Single-energy computed tomographic angiography.
Data are means ± 1 standard deviation.
Value was significantly lower (p < 0.017) than those with DECTA-40% and DECTA-80% groups.
Value was significantly lower (p < 0.017) than those with DECTA-80% group.
Value was significantly greater (p < 0.017) than those with DECTA-40% and DECTA-80% groups.
Table 5.
Arterial depiction on volume-rendered images
| Artery | SECTA | DECTA-40% | DECTA-80% | p-value |
|---|---|---|---|---|
| Brachiocephalic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Common carotid | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Subclavian | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Bronchial | 1.0 ± 0.0 | 1.0 ± 0.0 | 1.0 ± 0.0 | 1.00 |
| Internal thoracic | 1.1 ± 0.5 | 1.0 ± 0.0 | 1.0 ± 0.0 | 0.33 |
| Intercostal | 1.9 ± 1.2 | 1.3 ± 0.7 | 1.3 ± 0.5 | 0.087 |
| Common hepatic | 5.0 ± 0.0 | 4.9 ± 0.1 | 4.9 ± 0.3 | 0.99 |
| Proper hepatic | 5.0 ± 0.0 | 4.9 ± 0.1 | 4.9 ± 0.3 | 0.64 |
| Splenic | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Left gastric | 4.7 ± 0.8 | 3.9 ± 1.6 | 3.9 ± 1.6 | 0.035a |
| Gastroduodenal | 4.9 ± 0.5 | 4.7 ± 0.6 | 4.6 ± 0.8 | 0.10 |
| Inferior phrenic | 1.8 ± 1.2b | 1.1 ± 0.3 | 1.1 ± 0.3 | <0.001 |
| Superior mesenteric | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Inferior mesenteric | 2.3 ± 1.8 | 1.8 ± 1.6 | 1.8 ± 1.6 | 0.55 |
| Renal | 4.9 ± 0.4 | 4.9 ± 0.2 | 4.9 ± 0.2 | 0.70 |
| Lumbar | 3.0 ± 1.5b | 2.0 ± 1.4 | 2.1 ± 1.4 | 0.0066 |
| Common iliac | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| External iliac | 5.0 ± 0.0 | 5.0 ± 0.0 | 5.0 ± 0.0 | 1.00 |
| Internal iliac | 5.0 ± 0.0 | 4.6 ± 1.2 | 4.6 ± 1.2 | 0.10 |
| Iliolumbar | 4.9 ± 0.2b | 2.9 ± 1.6 | 2.9 ± 1.6 | <0.001 |
| Superior gluteal | 4.8 ± 0.4b | 4.0 ± 1.2 | 4.0 ± 1.2 | 0.0016 |
| Inferior gluteal | 4.9 ± 0.4c | 4.2 ± 1.2 | 4.4 ± 1.2 | 0.017 |
| Obturator | 3.6 ± 1.6 | 2.9 ± 1.9 | 2.9 ± 1.9 | 0.21 |
| Inferior epigastric | 3.6 ± 1.2 | 3.4 ± 1.4 | 3.4 ± 1.5 | 0.79 |
DECTA, Dual-energy computed tomographyic angiography; SECTA, Single-energy computed tomographic angiography.
Data are means ± 1 standard deviation.
Value was significantly different among three groups (p < 0.05), but there was no significant difference in pairwise comparisons.
Value was significantly greater (p < 0.017) than those with DECTA-40% and DECTA-80% groups.
Value was significantly greater (p < 0.017) than those with DECTA-40% group.
Figure 1.
A 78-year-old male with abdominal aortic aneurysm. Single-energyCTA obtained with standard iodine dose (600 mgI/kg) at 120 kVp. Anterior whole-body volume-rendered CTA (a) with and (b) without bone, and (c) maximum intensity projection image clearly shows the aorta, abdominal aortic aneurysm, and arterial branches, especially in intrapelvic arteries compared to dual-energy CTA in Figures 2 and 3. (d) Axial CTA of the pelvis clearly shows intrapelvic arteries. CTA, CT angiography.
Figure 2.
An 82-year-old male with aortic dissection in the ascending aorta. Dual-energy CTA obtained with 50% reduced iodine dose (300 mgI/ kg) at 40 keV. Anterior whole-body volume-rendered CTA reconstructed with ASiR of 40% (a) with and (b) without bone, and (c) maximum intensity projection image clearly shows the aorta and its arterial branches. (d) Axial CTA of the pelvis clearly shows the intrapelvic arteries, but background noise is relatively marked. ASiR, adaptive statistical iterative reconstruction; CTA, CT angiography.
Figure 3.
An 82-year-old male with aortic dissection in the ascending aorta, the same case as in Figure 2. Dual-energy CTA obtained with 50% reduced iodine dose (300 mgI/kg) at 40 keV. Anterior whole-body volume-rendered CTA reconstructed with ASiR of 80% (a) with and (b) without bone, and (c) maximum intensity projection image clearly shows the aorta and its arterial branches. (d) Axial CTA of the pelvis clearly shows the intrapelvic arteries, and background noise is reduced compared with the case shown in Figure 2d. ASiR, adaptive statistical iterative reconstruction; CTA, CT angiography.
Image quality
The granulation, noise, and diagnostic acceptability in VR images were significantly better in the SECTA group than in the DECTA groups (p < 0.001–0.0061). For noise in axial images, a significant difference was found using the Kruskal–Wallis test (p = 0.0012), but there was no significant difference when pairwise Mann–Whitney U test was conducted (p > 0.017). No significant difference was detected in the other parameters (p = 0.059–0.72) (Table 6).
Table 6.
Sharpness, granulation, noise, and diagnostic acceptability on CT angiography among three protocols
| Parameter | SECTA | DECTA-40% | DECTA-80% | p-value |
|---|---|---|---|---|
| Axial CTA | ||||
| Sharpness | 4.9 ± 0.2 | 4.9 ± 0.3 | 4.9 ± 0.2 | 0.47 |
| Granulation | 4.7 ± 0.4 | 4.5 ± 0.5 | 4.7 ± 0.5 | 0.19 |
| Noise | 4.6 ± 0.6 | 4.9 ± 0.3 | 5.0 ± 0.0 | 0.0012a |
| Diagnostic acceptability | 4.9 ± 0.3 | 4.9 ± 0.3 | 4.9 ± 0.2 | 0.38 |
| Mulitplanar-reformatted CTA | ||||
| Sharpness | 4.9 ± 0.2 | 4.9 ± 0.3 | 4.9 ± 0.2 | 0.60 |
| Granulation | 4.6 ± 0.5 | 4.5 ± 0.5 | 4.6 ± 0.5 | 0.72 |
| Noise | 4.8 ± 0.4 | 4.9 ± 0.3 | 5.0 ± 0.0 | 0.059 |
| Diagnostic acceptability | 4.9 ± 0.2 | 4.9 ± 0.3 | 4.9 ± 0.2 | 0.62 |
| Volume-rendered CTA | ||||
| Sharpness | 4.9 ± 0.2 | 4.8 ± 0.4 | 4.9 ± 0.4 | 0.18 |
| Granulation | 4.8 ± 0.4b | 4.5 ± 0.6 | 4.5 ± 0.5 | 0.0061 |
| Noise | 4.6 ± 0.6b | 3.1 ± 1.1 | 3.6 ± 1.0 | <0.001 |
| Diagnostic acceptability | 4.9 ± 0.2b | 3.6 ± 1.0 | 3.9 ± 0.9 | <0.001 |
DECTA, Dual-energy computed tomographic angiography; SECTA, Single-energy computed tomographic angiography.
Data are means ± 1 standard deviation.
Value was significantly different among three groups (p < 0.05), but there was no significant difference in pairwise comparisons.
Value was significantly greater (p < 0.017) than those with DECTA-40% and DECTA-80% groups.
Radiation exposure
The mean CTDIvol and DLP for the thorax, abdomen, and pelvis are summarized in Table 7. The mean CTDIvol and DLP for the thorax and abdomen were found to be significantly greater in the DECTA group than in the SECTA group (p < 0.0001–0.0058). Between the SECTA and DECTA groups, no significant difference was found in the mean CTDIvol and DLP for the pelvis (p = 0.054–0.055).
Table 7.
Radiation exposure between single- and dual-energy CT scans
| Parameter | SECTA | DECTA | p-value |
|---|---|---|---|
| CTDIvol (mGy) | |||
| Thorax | 16.9 ± 4.9 (7.1–28.2) | 23.3 ± 3.9 (11.4–25.1) | <0.0001a |
| Abdomen | 19.5 ± 5.6 (6.4–28.2) | 23.3 ± 3.9 (11.4–25.1) | 0.0031a |
| Pelvis | 21.3 ± 4.2 (10.3–27.4) | 23.3 ± 3.9 (11.4–25.1) | 0.055 |
| DLP (mGy*cm) | |||
| Thorax | 559.9 ± 171.4 (226.6–816.8) | 759.5 ± 152.5 (346.9–1020.6) | <0.0001a |
| Abdomen | 325.2 ± 100.3 (105.4–493.2) | 390.6 ± 75.5 (181.6–482.3) | 0.0058a |
| Pelvis | 487.5 ± 100.8 (235.1–660.8) | 538.4 ± 104.3 (241.5–662.1) | 0.054 |
CTDIvol, CT dose index volume; DECTA, Dual-energy computed tomographic angiography; DLP, Dose-length product; SECTA, Single-energy computed tomographic angiography.
Data are means ± 1 standard deviation with ranges in parentheses.
P < 0.05, significant difference.
Discussion
In this study, we compared quantitative parameters, arterial depiction, image quality, and radiation dose among the SECTA group at 120 kVp with standard iodine dose and DECTA groups at 40 keV with 50% reduced iodine dose. Our results demonstrated that the DECTA-80% group showed a slightly higher background noise, but higher CT number of the aorta and SNR while maintaining the arterial depiction in almost all arteries compared with SECTA. A previous study has reported that DECTA at 50 keV with 50% reduced iodine dose achieved 60% higher CT number of the aorta than SECTA at 120 kVp with standard iodine dose.13 In relation to this finding, the mean CT number of the aorta was increased by 58–68% in the DECTA-80% group. Although the mean background noise increased by 10–32%, the mean SNR could be increased by 23–59%.
The arterial depiction was almost comparable among SECTA, DECTA-40%, and DECTA-80% groups; however, that of the intrapelvic small arteries was recorded to be worse in both DECTA-40% and −80% groups than in the SECTA group. The increased background noise was determined as the cause of image degradation in the DE scan. At lower energy levels, VMIs are associated with increasing beam-hardening artifacts, often leading to worse image quality compared with conventional image reconstruction.14 These concerns regarding background noise and beam-hardening artifacts should be resolved in future generations of DECT.15 According to the previous results, the background noise in second-generation DECT was reduced by approximately half compared with that in first-generation DECT at 40–50 keV.16 Additionally, reduction of beam-hardening artifacts and improvement of spatial resolution at lower energy levels can be achieved by using a new-generation DECT.17 Moreover, the deep learning image reconstruction technique has been also proposed to reduce image noise and improve CT image quality in recent years.18,19 Therefore, we can expect that the image quality and arterial depiction of small or intrapelvic arteries will be improved by using noise reduction techniques in future generations of DECT.
In SECTA scanning, an automatic tube current modulation program was used. Through this program, the tube current is automatically adjusted based on regional body anatomies to maintain image noise and to improve radiation dose efficiency. Conversely, the CTDIvol was fixed at 10.68, 15.02, or 23.56 mGy on DE scans depending on the body size of the patient because automatic tube current modulation was unavailable on DE scans. This could lead to increased radiation dose in DE scans.20 Between the SECTA and DECTA groups in this study, the pelvic region was the only area where there was no statistical significant difference in radiation doses. Therefore, we believe that even in the DECTA protocol, the influence of the radiation dose on the gonad is small. Generally, the radiation dose is greater in the pelvic region compared with the abdominal region. In other words, an adequate amount of radiation dose is necessary in the pelvic region to maintain image quality. Therefore, we wonder if radiation dose was relatively insufficient in the pelvic region in DE scans. Recently, however, a new X-ray tube which is able to achieve high-dose radiation exposure of 1300 mA has been produced. We are expecting sufficient radiation dose in the pelvic region will be achieved with a new technology.
In this study, the mean patient body weight and body mass index were 63.7 kg (range, 39–95 kg) and 24.2 kg/m2 (range, 16.8–36.2 kg/m2), respectively. This body size distribution might be smaller than that of a western population. Increasing body mass index and abdominal fat significantly increases the radiation dose in abdominopelvic CT scans.21 This means that image degradation in the pelvic region might be more marked in a western population compared to an Asian population.
Our study had several limitations. First, the sample size was relatively small, which might have caused a selection bias. Second, the mean patient body weight and body mass index in this study might be smaller than that of a western population. Third, we could not directly compare the arterial depiction because we did not recruit patients who had completed the single- and dual-energy scans. Finally, we only used a first-generation fast kilovoltage-switching DECT scanner from a single vendor. Therefore, further clinical studies are needed to validate our results for other DECT scanners, such as dual-source or multilayer spectral DECT.
In conclusion, DECTA at 40 keV with adaptive statistical iterative reconstruction could reduce iodine dose by 50%, and in our study, this resulted in significantly higher vascular CT number and SNR and almost comparable arterial depiction. However, compared with SECTA, DECTA had a disadvantage of slightly increased radiation dose and degradation of arterial depiction in intrapelvic small arteries.
Contributor Information
Yoshifumi Noda, Email: noda1031@gifu-u.ac.jp.
Fumihiko Nakamura, Email: rad3132@gifu-u.ac.jp.
Noriyuki Yasuda, Email: yasunori@gifu-u.ac.jp.
Toshiharu Miyoshi, Email: t-miyosi@gifu-u.ac.jp.
Nobuyuki Kawai, Email: noburtcom@yahoo.co.jp.
Hiroshi Kawada, Email: h_kawada@gifu-u.ac.jp.
Fuminori Hyodo, Email: hyodof@gifu-u.ac.jp.
Masayuki Matsuo, Email: matsuo_m@gifu-u.ac.jp.
REFERENCES
- 1.Johnson PT, Heath DG, Kuszyk BS, Fishman EK. CT angiography with volume rendering: advantages and applications in splanchnic vascular imaging. Radiology 1996; 200: 564–8. doi: 10.1148/radiology.200.2.8685358 [DOI] [PubMed] [Google Scholar]
- 2.Leipsic J, Gurvitch R, Labounty TM, Min JK, Wood D, Johnson M, et al. Multidetector computed tomography in transcatheter aortic valve implantation. JACC Cardiovasc Imaging 2011; 4: 416–29. doi: 10.1016/j.jcmg.2011.01.014 [DOI] [PubMed] [Google Scholar]
- 3.Achenbach S, Delgado V, Hausleiter J, Schoenhagen P, Min JK, Leipsic JA. SCCT expert consensus document on computed tomography imaging before transcatheter aortic valve implantation (TAVI)/transcatheter aortic valve replacement (TAVR. J Cardiovasc Comput Tomogr 2012; 6: 366–80. doi: 10.1016/j.jcct.2012.11.002 [DOI] [PubMed] [Google Scholar]
- 4.Stavropoulos SW, Charagundla SR. Imaging techniques for detection and management of endoleaks after endovascular aortic aneurysm repair. Radiology 2007; 243: 641–55. doi: 10.1148/radiol.2433051649 [DOI] [PubMed] [Google Scholar]
- 5.Lehmkuhl L, Andres C, Lücke C, Foldyna B, Grothoff M, Scheinert D, et al. Dynamic CT angiography after abdominal aortic endovascular aneurysm repair: differences in contrast agent dynamics in the aorta and endoleaks--preliminary results. J Vasc Interv Radiol 2012; 23: 744–50. doi: 10.1016/j.jvir.2012.02.008 [DOI] [PubMed] [Google Scholar]
- 6.Hagiwara S, Saima S, Negishi K, Takeda R, Miyauchi N, Akiyama Y, et al. High incidence of renal failure in patients with aortic aneurysms. Nephrol Dial Transplant 2007; 22: 1361–8. doi: 10.1093/ndt/gfl779 [DOI] [PubMed] [Google Scholar]
- 7.Mita T, Arita T, Matsunaga N, Furukawa M, Zempo N, Esato K, et al. Complications of endovascular repair for thoracic and abdominal aortic aneurysm: an imaging spectrum. Radiographics 2000; 20: 1263–78. doi: 10.1148/radiographics.20.5.g00se161263 [DOI] [PubMed] [Google Scholar]
- 8.Shuman WP, O'Malley RB, Busey JM, Ramos MM, Koprowicz KM. Prospective comparison of dual-energy CT aortography using 70% reduced iodine dose versus single-energy CT aortography using standard iodine dose in the same patient. Abdom Radiol 2017; 42: 759–65. doi: 10.1007/s00261-016-1041-z [DOI] [PubMed] [Google Scholar]
- 9.Yu L, Leng S, McCollough CH. Dual-energy CT-based monochromatic imaging. AJR Am J Roentgenol 2012; 199(5 Suppl): S9–15. doi: 10.2214/AJR.12.9121 [DOI] [PubMed] [Google Scholar]
- 10.Matsumoto K, Jinzaki M, Tanami Y, Ueno A, Yamada M, Kuribayashi S. Virtual monochromatic spectral imaging with fast kilovoltage switching: improved image quality as compared with that obtained with conventional 120-kVp CT. Radiology 2011; 259: 257–62. doi: 10.1148/radiol.11100978 [DOI] [PubMed] [Google Scholar]
- 11.Kanematsu M, Goshima S, Miyoshi T, Kondo H, Watanabe H, Noda Y, et al. Whole-body CT angiography with low tube voltage and low-concentration contrast material to reduce radiation dose and iodine load. AJR Am J Roentgenol 2014; 202: W106–16. doi: 10.2214/AJR.13.10720 [DOI] [PubMed] [Google Scholar]
- 12.Noda Y, Goshima S, Kozaka K, Yoneda N, Mizuno N, Kato A, et al. Optimal window settings in single-source dual-energy computed tomography of the abdomen. Eur J Radiol 2018; 109: 204–9. doi: 10.1016/j.ejrad.2018.10.012 [DOI] [PubMed] [Google Scholar]
- 13.Shuman WP, Chan KT, Busey JM, Mitsumori LM, Koprowicz KM. Dual-energy CT Aortography with 50% reduced Iodine Dose Versus Single-energy CT Aortography with Standard Iodine Dose. Acad Radiol 2016; 23: 611–8. doi: 10.1016/j.acra.2015.12.019 [DOI] [PubMed] [Google Scholar]
- 14.Wu R, Watanabe Y, Satoh K, Liao Y-P, Takahashi H, Tanaka H, et al. Quantitative comparison of virtual monochromatic images of dual energy computed tomography systems: beam hardening artifact correction and variance in computed tomography numbers: a phantom study. J Comput Assist Tomogr 2018; 42: 648–54. doi: 10.1097/RCT.0000000000000726 [DOI] [PubMed] [Google Scholar]
- 15.Grant KL, Flohr TG, Krauss B, Sedlmair M, Thomas C, Schmidt B. Assessment of an advanced image-based technique to calculate virtual monoenergetic computed tomographic images from a dual-energy examination to improve contrast-to-noise ratio in examinations using iodinated contrast media. Invest Radiol 2014; 49: 586–92. doi: 10.1097/RLI.0000000000000060 [DOI] [PubMed] [Google Scholar]
- 16.Marin D, Ramirez-Giraldo JC, Gupta S, Fu W, Stinnett SS, Mileto A, et al. Effect of a noise-optimized second-generation monoenergetic algorithm on image noise and Conspicuity of hypervascular liver tumors: an in vitro and in vivo study. AJR Am J Roentgenol 2016; 206: 1222–32. doi: 10.2214/AJR.15.15512 [DOI] [PubMed] [Google Scholar]
- 17.Sugawara H, Takayanagi T, Ishikawa T, Katada Y, Fukui R, Yamamoto Y, et al. New fast kVp switching dual-energy CT: reduced severity of beam hardening artifacts and improved image quality in Reduced-Iodine virtual monochromatic imaging. Acad Radiol 2020; 27: 1586–93. doi: 10.1016/j.acra.2019.11.015 [DOI] [PubMed] [Google Scholar]
- 18.Akagi M, Nakamura Y, Higaki T, Narita K, Honda Y, Zhou J, et al. Deep learning reconstruction improves image quality of abdominal ultra-high-resolution CT. Eur Radiol 2019; 29: 6163–71. doi: 10.1007/s00330-019-06170-3 [DOI] [PubMed] [Google Scholar]
- 19.Singh R, Digumarthy SR, Muse VV, Kambadakone AR, Blake MA, Tabari A, et al. Image quality and lesion detection on deep learning reconstruction and iterative reconstruction of Submillisievert chest and abdominal CT. AJR Am J Roentgenol 2020; 214: 566–73. doi: 10.2214/AJR.19.21809 [DOI] [PubMed] [Google Scholar]
- 20.Sugawara H, Suzuki S, Katada Y, Ishikawa T, Fukui R, Yamamoto Y, et al. Comparison of full-iodine conventional CT and half-iodine virtual monochromatic imaging: advantages and disadvantages. Eur Radiol 2019; 29: 1400–7. doi: 10.1007/s00330-018-5724-4 [DOI] [PubMed] [Google Scholar]
- 21.Chan VO, McDermott S, Buckley O, Allen S, Casey M, O'Laoide R, et al. The relationship of body mass index and abdominal fat on the radiation dose received during routine computed tomographic imaging of the abdomen and pelvis. Can Assoc Radiol J 2012; 63: 260–6. doi: 10.1016/j.carj.2011.02.006 [DOI] [PubMed] [Google Scholar]