Atherosclerosis, which usually affects older individuals, induces vessel stenosis, occlusion, or arterial aneurysm.1 Computed tomography angiography (CTA) is the preferred method for imaging examination and follow-up of vasculopathy.2,3 In cervicocerebral CTA, the higher iodinated contrast agent dose and injection rate increased the risk of contrast agent–related adverse effects.4 After application of iodinated contrast agent, older patients often develop contrast-induced nephropathy because of decreased renal blood flow and compensatory capacity.5 The onset of contrast-induced nephropathy not only induces longer hospital stays and higher cost, but also increases morbidity and mortality of the related adverse events.6 Therefore, contrast agent dose should be reduced enough to satisfy the diagnostic criteria.
Currently, for contrast agent dosing in cervicocerebral CTA, weight-based contrast injection is recommended.7,8 This considers the impact of body weight on vascular enhancement, which is convenient and available in clinical practice. However, other factors related to contrast agent dosing demonstrate significant interindividual differences, such as cardiac output (cardiovascular circulation time), body mass index (BMI), body surface area (blood volume), and fat mass.9,10 Hence, determining contrast dosing solely based on weight is not conducive to accurate calculation of personalized contrast volume, particularly in children and obese patients. In obese patients, contrast dose might be overestimated because fat has little effect on dilution and dissemination of the contrast agent.11 For patients with a large weight and BMI but rapid circulation, calculating contrast dose based on body weight might cause ineffective contrast injection. In such cases, ineffective contrast injection results in wasting of the contrast agent and increased burden on the kidney. Moreover, the artifacts induced by retention of contrast agent impair the observations of the opening and initial segment of the ipsilateral vertebral artery and arteria subclavia. For patients with small weight and BMI but slow circulation, the results of arteriography might be unsatisfactory because the optimum peak duration is missed.
To improve image quality and reduce contrast dose, we adjusted the doses of the contrast agent and normal saline used in cervicocerebral CTA for older patients with normal cardiac function and BMI (50 kg ≤ weight < 80 kg) individually, according to the time to peak and number of time points to peak (cardiovascular circulation time) in the test bolus technique. This study might provide a novel insight for optimizing image quality and reducing contrast dose.
MATERIALS AND METHODS
Overall, 255 patients, who underwent cervicocerebral CTA in our hospital, were prospectively included in this study from July 2020 to May 2021. Inclusion criteria were as follows: 60 years ≤ age < 90 years, 18.5 kg/m2 ≤ BMI < 25 kg/m2, 50 kg ≤ weight < 80 kg; heart rate 60–100 beats/min, systolic pressure <140 mm Hg, and diastolic pressure <90 mm Hg. Patients with any of the following conditions were excluded from this study: (1) allergy to iodinated contrast agent; (2) cardiac, hepatic, or renal insufficiency, liver cirrhosis, portal hypertension, status asthmaticus, hyperthyroidism, myasthenia gravis, pheochromocytoma, hyperproteinemia, or multiple myeloma; (3) uncooperative with the examination, dyscrasia, severe or critically illness, patients in the perioperative period; (4) stent implantation, aneurysm clipping, embolization, or a significant motion artifact that impaired image observation; (5) number of time points to peak <4 or >9; (6) pallidum calcification; (7) lesion-induced increase or decrease in blood circulation rate; and (8) lesion-induced increase or decrease in myocardial contractility or history of long-term drug use. Accordingly, 21 patients were excluded from this study because of allergy to iodinated contrast agent (n = 2), significant motion artifact (n = 3), aneurysm clipping (n = 1), embolization (n = 1), stent implantation (n = 2), severe cervicocerebral stenosis (n = 3), pulmonary arterial hypertension (n = 2), renal insufficiency (n = 3), and heart disease (n = 4) (Fig. 1). Finally, 234 patients (144 male and 90 female patients with age range of 60–89 years and average age of 73 years) were included in this study. Patients were divided into groups A (n = 110) and B (n = 124) before undergoing scanning. All the experiments were approved by the ethics committee of Hunan Provincial People's Hospital (The First Affiliated Hospital of Hunan Normal University), and all subjects were informed of the purpose of this study and provided their informed consent.
The CTA scan in this study was carried out by using the following equipment/materials: Lightspeed 64-row 128-slice VCT (GE Healthcare, Chicago, IL), ADW4.7 postprocessing workstation (GE Healthcare), MEDRAD double-barrel high-pressure syringe (Bayer, Leverkusen, Germany), and iodixanol solution (320 mg I/mL; Jiangsu Hengrui Pharma Co, Ltd, Lianyungang, China). The patients had to remove metal foreign bodies prior to the scan (such as false teeth), had to be cooperative, and were not allowed to move during the scan. The scan ranged from 1 to 2 cm below the bifurcation of the trachea to the calvarium. Patients in groups A and B were injected with the contrast agent in the right median cubital vein via an 18-gauge catheter. The injection rates of contrast agent and normal saline were 4.5 mL/s. Before the injection, the contrast agent was heated to 37°C, and the contrast agent was fully hydrated within the body before and after the injection.
Test Bolus Protocol
The protocols for groups A and B were the same. Following injection of 15 mL iodixanol, 20 mL normal saline was infused at the same rate. Dynamic axial CT scans of the arcus aortae were performed at the same layer using the following parameters: scan field of view of 32 cm, matrix of 512 × 512, section thickness of 5 mm, scan time of 1.0 seconds, and scan interval of 1.0 seconds (each scan represented 2 seconds). Following consecutive scans for 6 to 12 time phases (the scan was stopped when the density began to decrease after an initial increase), MROI analysis software was used to plot the time-density curve of the enhanced arcus aortae turning point (the time with the highest CT value). Time to peak and scan delay time for CTA were determined with the following formulas: time to peak (in seconds) = n × 2 + 8 seconds (n represented the number of time points to peak); delay time (in seconds) = time to peak (in seconds) − plain scan time (in seconds). The region of interest (ROI) was placed in the center of the lumen cross-section to avoid soft plaque, stenosis areas, and artifacts.
The CTA sequence included plain scans, which was immediately followed by enhanced scans. The scan modes of groups A and B were spiral scans (tube voltage of 120 kVp, automatic tube current modulation of 100 to 650 mA, layer thickness/distance of 0.625 mm, field of view of 320 mm, matrix of 512 × 512, noise index of 10.0 HU, tube rotation speed of 0.4 second per rotation, collimator width of 40 mm, and pitch of 0.984). The doses of contrast agent (1.0 mL/kg) and normal saline (40 mL) in group A were determined in reference to previous studies.7,8 For group B, the doses of contrast agent and normal saline were adjusted based on the test bolus time to peak and the number of time points to peak. The contrast dose was calculated as follows: contrast dose (in milliliters) = contrast injection duration (in seconds) × injection rate of contrast agent (in milliliters per second); contrast injection duration (in seconds) = time to peak (in seconds) − normal saline injection duration (in seconds). Normal saline was injected at a predefined dose. Specifically, 30 mL normal saline was injected when the number of time points to peak was 4 to 5, with an injection duration of 7 seconds; 40 mL normal saline was injected when the number of time points to peak was 6 to 7, with an injection duration of 9 seconds; and 50 mL normal saline was injected when the number of time points to peak was 8 to 9, with an injection duration of 11 seconds. The starting and ending points of plain and enhanced scans and the position of the tube were kept the same to allow CTA subtraction. The contrast agent was injected at the beginning of the plain scan. The obtained data were uploaded to a GE ADW4.7 workstation, and then multiplanar reformation, volume reformation, curved planar reformation, and maximum intensity projection in the 2 groups were conducted using the same parameters.
On the thin-slice, enhanced axial images, the CT values of the arcus aortae, the bifurcations of left and right common carotid arteries, and the left and right middle cerebral artery M1 segments (CT vessel) were evaluated. Their SDs were defined as noise (SD vessel). In addition, the CT values of the left pectoralis major, left and right sternocleidomastoids, and left and right lenticular nuclei were assessed as background CT values (CT background). The signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated as follows: SNR = CT vessel/SD vessel; CNR = (CT vessel − CT background) / SD vessel. The CT value of the right transverse sinus was measured to evaluate the enhancement of the venous sinus. The ROI was placed in the center of the tissues to avoid skeleton, arterial plaque, cerebrospinal fluid, and artifacts. To measure the CT values of the vessels, the size of the ROI was chosen to be larger than half of the diameter of the lumen and consistent in the left and right sides. Each site was scanned twice, and the CT values were then averaged.
The image quality was independently evaluated by 2 radiologists who were blinded to the study and had more than 10 years of work experience. The quality of the original and reconstructed images was subjectively scored on a 5-point scale.12 A score of 1 (Fig. 2A) indicated poor image contrast, coarse image particles, rough vascular margin, and inability to display the trunk and branches of the cervicocerebral vessels, with severe artifacts that affected the diagnostic accuracy. A score of 2 (Fig. 2B) signified poor image contrast, coarse image particles, rough vascular margin, and blurring of the main branches and distal ends of the cervicocerebral vessels with significant artifacts that affected the diagnostic accuracy. A score of 3 (Fig. 2C) signified substandard image contrast, nonuniform image particles, a clear display of the trunk and main branches of the vessels, inability to display the branches and distal ends, and no effect on clinical diagnosis. A score of 4 (Fig. 2D) signified relatively good image contrast; uniform image particles; clear display of the trunk, branches, and distal ends of the vessels with few artifacts; and no effect on diagnostic accuracy. A score of 5 (Fig. 2E) indicated excellent image contrast, uniform image particles, and continuous vessels with a sharp margin that clearly showed the trunk, branches, and distal ends of the vessels without any artifact and satisfied the clinical diagnostic requirements. Images with a score of ≥3 met clinical diagnostic requirements, images with a score of 4 were considered good diagnosis, and a score of 5 represented an excellent image. Contrast retention and artifacts in the right subclavian vein, right brachiocephalic veins, and superior vena cava were evaluated using a 4-grade system.13 Grade 1 represented no retention of the contrast agent of high concentration in the veins. Grade 2 indicated laminar contrast only in the posterior vena cava lumen, slight beam hardening artifacts, and no effect on arteriography. Grade 3 represented contrast retention throughout the veins, with medium beam hardening artifacts, with clinical diagnostic requirements satisfied. Lastly, grade 4 indicated contrast retention throughout the veins, with severe beam hardening artifacts that affected the local diagnostic accuracy.
Experimental data were collated and analyzed using Excel (Microsoft Corp, Redmond, WA) and SPSS 19.0 (IBM SPSS Inc, Armonk, NY). For quantitative data with a normal distribution, statistical significance was evaluated using the unpaired t test, whereas χ2 and t tests were used to analyze qualitative data. For ranked data and quantitative data without a normal distribution, the statistical significance was assessed using the rank-sum test. A κ test was used to evaluate the consistency of subjective scores of image quality made by the 2 radiologists (κ < 0.40: poor; 0.40 ≤ κ < 0.75: medium; κ ≥ 0.75: good). P < 0.05 represented statistical significance.
Comparisons of Patient Characteristics and Contrast Agent Dose
There was no significant difference in sex, age, body weight, height, BMI, heart rate, blood pressure, time to peak, and number of time points to peak between groups A and B (P > 0.05), whereas the contrast agent dose used in group B was decreased by 18.2% relative to group A (P < 0.05) (Table 1).
TABLE 1 -
Comparisons of Clinicopathological Features, Time to Peak, Number of Time Points to Peak, and Contrast Dose in the 2 Groups
(n = 110)
(n = 124)
|Age, mean (SD), y
|Gender, n (%)
|Height, mean (SD), cm
|Weight, mean (SD), kg
|BMI, mean (SD), kg/m2
|HR, mean (SD), beats/min
|Blood pressure, mean (SD), mm Hg
|No. time points to peak
|Time to peak, mean (SD), s
|Contrast dose, mean (SD), mL
Objective Evaluation of Image Quality
The CT values of the arcus aortae, bifurcations of left and right common carotid arteries, and left and right middle cerebral artery M1 segments for patients in groups A and B were higher than 300 HU, which met the imaging diagnostic requirements for cervicocerebral vascular disorders. The CT values of arteries showed no significant differences between groups A and B (P > 0.05). However, compared with group A, the noise at the bifurcation of the right common carotid artery in group B increased by 1.7%, and the SNR of the left middle cerebral artery M1 segment in group B decreased by 6.6%. The noise, CNR, and SNR of other arteries of interest showed no statistically significant difference between the 2 groups (P > 0.05). Moreover, groups A and B showed nonsignificant differences in the CT value of the transverse sinus (P > 0.05) (Table 2).
TABLE 2 -
Comparison of Image Quality in the 2 Groups
||Group A (n = 110)
||Group B (n = l24)
| Subjective score of image quality of original and reconstructed images (score)
| Contrast retention and artifacts in right subclavian vein, right brachiocephalic veins, and superior vena cava (grade)
| CT values in transverse sinus, mean (SD), HU
| Arcus aortae, mean (SD)
| Bifurcation of left common carotid artery, mean (SD)
| Bifurcation of right common carotid artery, mean (SD)
| Left middle cerebral artery M1 segment, mean (SD)
| Right middle cerebral artery M1 segment, mean (SD)
DBP, diastolic blood pressure; SBP, systolic blood pressure.
Subjective Evaluation of Image Quality
Subjective evaluation of image quality by the 2 radiologists showed good consistency (κ = 0.815). All the original and reconstructed images had sufficient quality for use in clinical diagnosis. Specifically, group A consisted of 71 cases with a score of 5 (Figs. 3A–F), 36 cases with a score of 4, and 3 cases with a score of 3. Group B consisted of 87 cases with a score of 5 (Figs. 4A–F), 34 cases with a score of 4, and 3 cases with a score of 3. There was no significant difference in the subjective scores of image quality between groups A and B (P > 0.05). Further, we compared the contrast retention and artifacts in the right subclavian vein, right brachiocephalic veins, and superior vena cava in groups A and B. The results suggested that the number of grades 1 and 2 images increased by 23.4% in group B relative to group A. These results demonstrated reduced contrast retention and artifacts in group B as compared with group A (Table 2, P < 0.05).
In the present study, we adjusted the doses of contrast agent and normal saline, based on time to peak and number of time points to peak in the test bolus scan in older patients who underwent cervicocerebral CTA. Our findings demonstrated that the CTA image quality satisfied diagnostic requirements, reduced contrast retention in the veins on the injection side, and decreased the contrast dose required.
Arterial enhancement could be affected by multiple factors, such as CT parameters, patient-related factors, and contrast agent–related factors (volume, injection rate, injection duration, bolus injection, and saline flush).9,14,15 Computed tomography parameters affecting contrast enhancement include delay time, scan duration, tube voltage, and scan speed and direction.16 Body weight, BMI, body surface area (blood volume), and cardiac output (cardiovascular circulation time) are significant patient-related factors involved in vascular CT contrast enhancement.17 Other factors, such as age, sex, transvenous pathway, renal function, liver cirrhosis, and portal hypertension, were considered as patient-related factors that could influence contrast enhancement.
Peak and time to peak of arterial enhancement are mainly affected by cardiac output.18,19 Arterial enhancement is directly proportional to the iodine content in blood vessels, which is related to iodine input and hemodilution.20 Patients with longer time to peak have lower peak arterial enhancement, and concomitantly, more intravascular contrast agent is dispersed and diluted because of the blood circulation. When the concentration and injection rate of contrast agent remain unchanged, prolonging the injection duration can increase the intravascular iodine input. However, it also broadens the bolus injection and enhances the cumulative effect of recirculation,21,22 so that arterial enhancement is increased uniformly, and distal small blood vessel filling is improved. The test bolus technique could accurately capture individual time to peak and delay time.23 A cycle through the cervicocerebral arteries and veins is approximately 6 to 8 seconds. At present, CT machines with 64 or more rows have a rotation speed of ≤0.4 second per rotation and can complete the scan within 1 cervicocerebral cycle (scan time ≤4 seconds). The time to peak of the research group was postulated to be equal to the injection duration of the contrast agent and normal saline. For patients with a longer time to peak, the injection duration of the normal saline was extended to obtain a reasonable injection duration and contrast agent dose. The scan started at the peak when all the contrast agent and normal saline had been injected. This ensured arterial enhancement, but also reduced contrast retention in the veins of the injection side. Patients in the 2 groups were scanned at the same tube voltage and current, and there was no statistically significant difference in sex, age, height, weight, BMI, heart rate, blood pressure, time to peak, and number of time points to peak between the 2 groups. Therefore, there was no substantial change in the intra-arterial CT values between the 2 groups, and the noise, SNR, and CNR between the 2 groups showed only slight differences. Because of the moderate increase in the noise and the slight decreases in SNR and CNR, the arterial enhancement was hardly affected. Thus, the subjective evaluation of image quality was not significantly different. Moreover, there was no statistically significant difference in the CT values of the transverse sinuses between the 2 groups, which may be attributed to the longer contrast injection duration and peak duration, resulting in venous/venous sinus visualization. This may also be related to interindividual differences in vascular enhancement. A future study should investigate how to reduce venous/venous sinus enhancement in cervicocerebral CTA examination.
Moreover, among the 234 patients, patients with 5, 6, and 7 time points to peak accounted for 78.2% (183 cases), and those with 4, 8, and 9 time points to peak accounted for 11.5% (27 cases), 7.3% (17 cases), and 3.0% (7 cases), respectively. In addition, we also found that the image quality in patients with 5, 6, and 7 time points to peak in group B was better than that of patients with 4 or 9 time points to peak, as indicated by higher subjective scores of image quality, less cephalic vein/venous sinus contamination, and increased SNR and CNR. The number of patients with 4, 5, 6, and 7 time points to peak in group B accounted for 90.0% (112 cases). Overall, the contrast agent dose in group B was reduced as compared with group A. Previously, Hubbard et al24 attempted to dilute the contrast agent in the test bolus technique and demonstrated that use of a diluted test bolus combined with CTA improved aortic and coronary enhancement and the CNR. In the present study, we adjusted the contrast dose based on time to peak and number of time points to peak in test bolus, which could reduce the contrast dose based on interindividual differences.
This study had some limitations. First, for ethical and moral issues, we did not perform intraindividual comparisons of the 2 protocols of contrast and normal saline dosing. Second, because of the small number of cases with time points to peak of less than 4 or more than 9, the patients in this study were not stratified according to the number of the time points to peak. Third, because of the difficulty of obtaining digital subtraction angiography data for the same individual during the same period, detection of the lesions in the 2 groups was not compared with the digital subtraction angiography criterion standard. Last, because there are many factors affecting arterial enhancement, patients with 18.5 kg/m2 > BMI ≥ 25 kg/m2, 50 kg > weight ≥ 80 kg, or abnormal cardiac function were not included in the study. More in-depth investigations will be required to elucidate these issues.
In conclusion, for older patients with normal cardiac function and BMI (50 kg ≤ weight < 80 kg), adjusting the doses of the contrast agent and normal saline in cervicocerebral CTA based on time to peak and number of time points to peak with the test bolus could reduce the contrast dose and decrease contrast retention and artifacts in the veins on the injection side, while satisfying the diagnostic image quality requirements.
The authors thank Zhang Wentao, Peng, Qi, Tang Bei, Wang Shisheng, and Jiang Ziling (CTA scan); He Rong and Zhou Jinfeng (patient collection); Ren Yong and Che Yilei (statistical analysis); Liu Qingxiong and Tian Min (design of CTA protocol and image evaluation).
1. Haidegger M, Kneihsl M, Niederkorn K, et al. Blood biomarkers of progressive atherosclerosis and restenosis after stenting of symptomatic intracranial artery stenosis. Sci Rep
2. Choi JH, Jang J, Koo J, et al. Multiphasic computed tomography angiography findings for identifying pseudo-occlusion of the internal carotid artery. Stroke
3. Verdoia M, Gioscia R, Marcolongo M, et al. Noninvasive imaging risk stratification with computed tomography angiography for coronary artery disease. Cardiol Clin
4. Zhang F, Lu Z, Wang F. Advances in the pathogenesis and prevention of contrast-induced nephropathy. Life Sci
5. Huyut MA. Kidney injury molecule-1 is associated with contrast-induced nephropathy in elderly patients with non-STEMI. Arq Bras Cardiol
6. Vicente-Vicente L, Gonzalez-Calle D, Casanova AG, et al. Quercetin, a promising clinical candidate for the prevention of contrast-induced nephropathy. Int J Mol Sci
7. Association RSoCM. Expert consensus on head and neck CT angiography scanning scheme and injection scheme. Chin J Radiol
8. ACR-ASNR-SPR Practice Parameter for the Performance and Interpretation of Cervicocerebral Computed Tomography Angiography
. Reston, Virginia: ACR; 2020:1–14.
9. Bae KT. Intravenous contrast medium administration and scan timing at CT: considerations and approaches. Radiology
10. Kondo H, Kanematsu M, Goshima S, et al. Body size indexes for optimizing iodine dose for aortic and hepatic enhancement at multidetector CT: comparison of total body weight, lean body weight, and blood volume. Radiology
11. Kan J, Milne M. Aorta, liver, and portal vein CT contrast enhancement during the portal venous phase are positively associated with abdominal fat percentage in dogs. Vet Radiol Ultrasound
12. Cai W, Gong J, Hu C, et al. Feasibility study of full model iterative recombination technique combined with low voltage and low contrast iodine intake in CT angiography of head and neck. Chin J Radiol
13. Liang Y, Mo X, Deng D, et al. Study of dose optimizing of test-bolus in intracranial and cervical CTA: the effect of joint application of “three lows” and thinner collimator. J Clin Radiol
14. Bae KT, Heiken JP. Scan and contrast administration principles of MDCT. Eur Radiol
. 2005;15(Suppl 5):E46–E59.
15. Bae KT, Seeck BA, Hildebolt CF, et al. Contrast enhancement in cardiovascular MDCT: effect of body weight, height, body surface area, body mass index, and obesity. AJR Am J Roentgenol
16. Kalra MK, Becker HC, Enterline DS, et al. Contrast administration in CT: a patient-centric approach. J Am Coll Radiol
17. Nijhof WH, Hilbink M, Jager GJ, et al. A non-invasive cardiac output measurement as an alternative to the test bolus technique
during CT angiography. Clin Radiol
18. Fleischmann D. CT angiography: injection and acquisition technique. Radiol Clin North Am
. 2010;48:237–247, vii.
19. Nijhof WH, Jansen MM, Jager GJ, et al. Feasibility of a low concentration test bolus in CT angiography. Clin Radiol
20. Koudrina A, O’Brien J, Garcia R, et al. Assessment of aptamer-targeted contrast agents for monitoring of blood clots in computed tomography and fluoroscopy imaging. Bioconjug Chem
21. Fleischmann D. How to design injection protocols for multiple detector-row CT angiography (MDCTA). Eur Radiol
. 2005;15(Suppl 5):E60–E65.
22. Faggioni L, Gabelloni M. Iodine concentration and optimization in computed tomography angiography: current issues. Invest Radiol
23. Das K, Biswas S, Roughley S, et al. 3D CT cerebral angiography technique using a 320-detector machine with a time-density curve and low contrast medium volume: comparison with fixed time delay technique. Clin Radiol
24. Hubbard L, Malkasian S, Zhao Y, et al. Contrast-to-noise ratio optimization in coronary computed tomography angiography: validation in a swine model. Acad Radiol