Dual-source Computed Tomography in Paediatric Congenital Heart Disease Patients - Combination of Low-kilovoltage Protocols and Ultravist Injection

Login or register to view PDF.
ICR - Volume 3 Issue 1;2008:3(1):16-20

In the clinical management of patients with complex congenital heart disease (CHD), accurate 3D evaluation of their morphological conditions is critical. 3D imaging should demonstrate the shape and spatial relation of great arteries, proximal branch pulmonary arteries and anomalous pulmonary venous or systemic connections, and eventually coronary artery course. 3D information on extra-cardiac morphological characteristics may determine the choice and the route of surgical intervention.
Multislice CT has been proposed as a tool for 3D anatomical visualisation in patients with CHD, and is increasingly being used in many institutions.1 Multislice technology, including most recently dual-source computed tomography (DSCT), provides volume acquisition in a short period of time, making high-quality 3D vascular images possible even in neonates or infants. The multislice CT technology now available has a very short acquisition time, which drastically reduces respiratory artefacts. The effect of heart motion may be removed by synchronisation of the acquisition with the cardiac rhythm.
In our surgical centre, which specialises in CHD, multislice CT has rapidly become an important complementary imaging technique for both pre-and post-operative management of patients. The aim of this article is to present our initial experience of DSCT over a 18-month period, describing the protocols necessary for high-quality images using low radiation exposure settings.

Technical Aspects

DSCT allows very short acquisition time, sub-millimetre slice thickness and electrocardiogram (ECG)-gated acquisition – even in children with a high heart rate – due to high temporal resolution (83ms down to 42ms). The first question is whether one should use ECG-gated acquisition in CHD patients, and the second question is which protocol is most suitable. In particular, the radiation dose delivered should be estimated for each protocol in order to minimise radiation exposure.2

Neonates and Young Infants
Non-electrocardiogram-gated Acquisitions

There are three main advantages to the use of non-ECG-gated acquisitions in neonates and young infants:
• Images of extra-cardiac structures are less sensitive to heart motion than the heart itself. ECG gating may therefore be unnecessary for diagnostic purposes when the clinical picture involves extra-cardiac anatomy. • Respiratory artefacts are responsible for substantial degradation of images. These respiratory artefacts may cause more significant artefacts than heart motion does. Non-ECG-gated acquisitions are acquired more quickly than ECG-gated acquisitions, and thus are responsible for fewer respiratory artefacts. Using DSCT, the thorax of a neonate can be scanned in about one second, so respiratory artefacts are minimal. • ECG-gated cardiac retrospective acquisitions may require a much higher radiation dose (up to four times higher) than non-gated thoracic CT because the exposure time is longer due to low pitch acquisitions. In babies, organ sensitivity to radiation is much higher than in adults, and a risk of developing cancer in the future cannot be totally ruled out.3

Electrocardiogram-gated Acquisitions

However, when coronary artery visualisation is required, ECG-gated acquisition may be recommended to improve coronary delineation and image quality.4 In our experience, in practice this is applicable only using 64-slice CT or DSCT, the temporal resolution being insufficient with earlier technology. In our experience, the image quality of coronaries in neonates has been greatly improved with the move from 64-slice CT to DSCT due to much better temporal resolution (83 versus 165ms). This improvement has immediate clinical implications: for example, in our institution babies with tetralogy of Fallot used to have systematic coronary angiography to look for a possible coronary anomaly before surgical intervention; however, this procedure has now been totally replaced by coronary CT angiography, which allows reliable 3D depiction of the coronary course without the need for coronary angiography.

Infants Over Seven Years of Age and Adults

In this group of older infants or young adults, two options are possible: either breath-hold angio-CT acquisition or ECG-gated acquisition. In this group, the protocol must be chosen depending on the clinical picture. For example, if coronary visualisation is required to look for a possible anatomical variant, ECG-gated protocols are recommended. In other cases, ECG gating is non-mandatory. In any case, one should always keep in mind that the radiation dose delivered is much higher when using ECG-gated acquisition.

Dose Consideration

Radiation exposure is a major public health issue. CT contributes greatly to the population dose arising from medical exposure, accounting for 35% of the dose delivered during diagnostic examinations even though it represents only 4% of such studies. Multislice CT offers even more diagnostic capabilities but tends to increase the radiation dose due to routine use of thinner slice thickness, extension of the volume of acquisition or multiple-phase acquisitions. Following the ALARA principle (‘as low as reasonably achievable’), dose reduction is necessary, but examination quality must be preserved without losing diagnostic information.5,6
The thorax is a low-attenuation region, although substantial dose reduction during chest CT is feasible because of the high inherent contrast. In our centre, we decided to apply the ALARA principle as far as possible in neonates and babies with CHD, and then apply some systematic rules:
• systematic use of 80kV settings;7 • adaptation of the mAs to the child’s weight; and • only one-phase acquisition when possible.
Eighty kilovolt protocols have been successfully performed to scan the thorax of adults weighing less than 75kg without substantial loss of image quality.8 Reducing the kilovoltage from 120 to 80kV decreases the radiation dose to 65% of that produced at the constant current setting, as radiation dose varies with the square of kV. This setting is sufficient for high-quality images, as long as the mAs are adjusted according to the child’s weight. We use 80kV as the standard kilovoltage setting in children.7 Current exposure is adapted to bodyweight in neonates and infants: for non-ECG-gated acquisitions we recommend 10mAs per kg bodyweight up to 6kg.

Using this protocol, radiation exposure is estimated to be 1mSv for a neonate, which is equivalent to the dose delivered by natural radiation over a six-month period. Radiation dose due to CT acquisition may be lower than the radiation dose delivered during conventional angiography.9 Radiation dose associated with ECG-gated multislice CT is higher using the current retrospective mode: for example, in our institution, in a one-year-old baby thermo-luminescent measurements provide a dose level of 3.4mGy using ECG-gated acquisition. Coronary angiography in the same patient was associated with a very similar radiation level (3.1mGy). Anatomical data acquired from CT may be judiciously used to limit the number of views acquired using angiography, and can sometimes replace conventional angiography.10 CT may then be advantageous in terms of reducing global radiation exposure in CHD patients.
The other advantage of 80kV settings is the ability to reduce the amount of contrast medium injected, because a lower kilovoltage is more sensitive to contrast (iodine has a high atomic number) than the standard 120kV setting.

Injection Protocol

Dose injection must be adapted to the baby’s weight: in our institution, we currently use 2cc per kg of Iopromide (Ultravist) at a concentration of 300mg/ml. We did not record any serious adverse reactions over a seven-year period including more than 800 babies.
The basic protocol of injection of the pulmonary arteries or for systemic vascular enhancement is as follows. Using 80kV, the rate of injection can be as low as 0.5cc per second in neonates with a catheter placed in the vein of the hand. A higher rate may be used in cases where a central catheter (femoral or jugular) is used. A power injector is routinely used to ensure a continuous and regular flow rate. In cases of peripheral injection, the rate of injection varies from 0.5 to 1cc/s depending on the quality of venous access. The start delay in neonates and infants is 15 seconds for peripheral injection and 10 seconds for central venous injection. To be sure of having vascular contrast during the acquisition, we sometimes slightly increase the amount of contrast medium in order to follow this rule:
time of injection = start delay + time of acquisition
Using this rule, acquisition is never ‘too late’ for good vascular enhancement, because acquisition ends with the end of the injection, so the contrast medium is still in the peripheral veins when acquisition ends.

Precautions for Venous Access

Peripheric venous access is achieved in the paediatric unit. Right-arm injection is preferable (but not mandatory) to avoid possibly striking artefacts on the innominate left brachio-cephalic vein. In some cases, venous connections are congenitally different or surgically modified. It is important to have this information, when available, before the scan procedure as it may change the scan injection protocol. Venous visualisation may be realised at first pass, with a high concentration of contrast medium, or sometimes later, at the time of venous return. The optimal injection protocol depends on each particular venous anatomy.
Catheter permeability is checked before the injection. It is essential to avoid any air injection during the scan procedure; all bubbles should be removed when connecting the catheter to the power injector. Because many patients with CHD have right-to-left shunt, air injection through venous access could cause systemic air embolism, with possibly fatal consequences. Extravasation of contrast may occur, with an incidence of 1.4% in our centre in 2007. These rare complications were treated immediately without any consequence.


In our experience, general anaesthesia is never necessary. In neonates we do not use any sedative drugs. In infants, we recommend oral or intra-rectal sedation (or both) before the CT procedure to prevent baby agitation during the acquisition, which may be responsible for poor image quality and, sometimes, need for re-examination. Sedation is not always mandatory if the baby is quiet. Experienced technologists are necessary in the CT room for good management of the babies: precise knowledge of managing babies and a gentle attitude are of primary importance.
Our sedation protocol in infants includes intra-rectal midazom at a dose of 0.3mg/kg given 15 minutes before examination. Additional sedative drugs may be useful (hydroxyzine at a dose of 1mg/kg orally one hour before examination). With experienced technologists, the mean total examination time in the CT room is between 15 and 20 minutes. Qualified medical monitoring may sometimes be necessary during the examination, depending on the clinical condition of the baby. In all cases, oxygen saturation should be closely monitored.

Anatomical Assessment
Pulmonary Arteries

Pulmonary artery evaluation is required in pulmonary atresia with ventricular septal defect, tetralogy of Fallot, truncus arteriosus or suspicion of pulmonary sling.11 For pulmonary artery visualisation, we usually do not use ECG-gated acquisition. With DSCT, we usually use 0.6mm collimation and obtain a 1mm slice width with an increment of 0.5mm. High resolution is advantageous for pulmonary artery stenosis evaluation. For neonates or infants below seven years of age, start delay is either 15 seconds (peripheral venous access) or 10 seconds (central venous access). In older patients (over six years of age), we may use the bolus tracking technique for optimisation of start delay. For reconstructions, 3D images are currently performed using maximum-intensity projection, multiplanar reformations and volume-rendering techniques (see Figure 1).

Coronary Arteries

Anomalous coronary arteries are frequently associated with CHD. A frequent anomalous finding is a left coronary artery originating from the right coronary sinus, but many variants are possible. In patients with tetralogy of Fallot, detection of an anomalous origin of coronaries is especially important before surgery when a ventriculotomy is planned, as accidental lesions of the coronary artery crossing the right ventricle during intervention can be fatal.
If non-ECG-gated acquisition is insufficient to visualise coronary origins,12 in neonates we recommend ECG-gated acquisition using DSCT to improve coronary anatomy visualisation.4 Due to the very high heart rate in babies, the maximal temporal resolution is required: bi-segmental reconstructions may therefore be useful to increase temporal resolution up to 43ms. In older CHD patients, if the patient can hold his or her breath for a sufficient time, ECG-gated acquisition is the technique of choice. Free-motion artefact visualisation makes it possible to precisely evaluate the coronary artery tree. The heart rate may be lowered with the help of beta-blockers. The maximum setting of the tube current is also carefully selected and adapted to the patient’s anatomy.13,14 If axial and maximum intensity projection images are insufficient for diagnosis, the volume-rendering technique may provide comprehensive imaging of the coronary artery anomaly (see Figure 2).

Aorta and Collaterals

Evaluation of the aortic anatomy is essential in cases of aortic coarctation or for suspicion of aortic arch anomalies: complete or incomplete double aortic arches (see Figure 3), right aortic arches or cervical arches are very clearly seen with 3D CT images. In cases of pulmonary atresia with ventricular septal defect, the major aorto-pulmonary collateral arteries often originate from the start of the descending aorta; evaluation of the size and spatial relationship of these arteries is of primary importance for planning surgical intervention. Thin collimation (1mm slice thickness or lower) is of interest to obtain high-resolution aortic images, especially in cases of aortic coarctation for a better evaluation of vessel narrowing. For reconstructions, 3D images are currently performed using maximum-intensity projection, mutliplanar reformations and volume-rendering techniques.

Upper Airways

Compression of the central airways of vascular origin may be due to various situations: the most frequently observed are aortic arch anomalies, pulmonary artery sling, dilated pulmonary arteries or posteriorly displaced aorta after switch intervention. In 20 consecutive cases of vascular compression, we observed six cases of aortic arch anomalies (four double aortic arches, one incomplete vascular ring, one circling aorta), five cases of dilatation of the pulmonary arteries (one case of pulmonary valve agenesia and four cases of pulmonary hypertension associated with ventricular septal defect), three cases of posteriorly displaced aorta after intervention for transposition of great vessels, three cases of left main bronchus compression by descending aorta, one pulmonary artery sling (see Figure 4) and two compressions by a brachio-cephalic artery.15
In addition to maximum-intensity projection reconstructions, the volume-rendering technique is very effective at showing central airway narrowing. Using the latter technique, it is easy to colourise low-density airway structures and high-density vascular structures to obtain high-contrast images, making diagnosis of vascular compression easy (see Figure 4).

Anomalous Venous Return

CT is a very efficient imaging technique to detect pulmonary or systemic anomalous venous return. Both techniques allow 3D visualisation of these anomalies. 3D images can be used as a primary method for diagnosis. ECG-gated acquisitions are not necessary because venous structures are not sensitive to cardiac motion. Injection site and timing of acquisition must be chosen carefully, because the time for best opacification depends on venous drainage and any anomalous venous drainage may interfere with the optimal timing. An additional delayed acquisition may be necessary to opacify the whole venous system.


Multislice CT, especially DSCT angiography, thus represents an important additional low-invasive diagnostic tool for the evaluation of CHD. Using Iopromide at 300mg/ml and low mAs and kilovoltage settings, excellent image quality may be provided in a very safe fashion, providing relevant information for optimal surgical or interventional management of these very young patients. Ôûá

  1. Gilkeson RC, Ciancibello L, Zahka K, Pictorial essay. Multidetector CT evaluation of congenital heart disease in pediatric and adult patients, AJR Am J Roentgenol, 2003;180: 973–80.
    Crossref | PubMed
  2. Hollingsworth CL, Yoshizumi TT, Frush DP, et al., Pediatric cardiac-gated CT angiography: assessment of radiation dose, AJR Am J Roentgenol, 2007;189:12–18.
    Crossref | PubMed
  3. Brenner D, Elliston C, Hall E, Berdon W, Estimated risks of radiation-induced fatal cancer from pediatric CT, AJR Am J Roentgenol, 2001;176:289–96.
    Crossref | PubMed
  4. Tsai IC, Lee T, Chen MC, et al., Visualization of neonatal coronary arteries on multidetector row CT: ECG-gated versus non-ECG-gated technique, Pediatr Radiol, 2007;37(8):818–25.
    Crossref | PubMed
  5. Vock P, CT dose reduction in children, Eur Radiol, 2005;15: 2330–40.
    Crossref | PubMed
  6. Willis CE, Slovis TL, The ALARA concept in pediatric CR and DR: dose reduction in pediatric radiographic exams—a white paper conference, AJR Am J Roentgenol, 2005;184:373–4.
    Crossref | PubMed
  7. Paul JF, Abada HT, Sigal-Cinqualbre A, Should low-kilovoltage chest CT protocols be the rule for pediatric patients?, AJR Am J Roentgenol, 2004;183:1172, author reply 1172.
    Crossref | PubMed
  8. Sigal-Cinqualbre AB, Hennequin R, Abada HT, et al., Lowkilovoltage multi-detector row chest CT in adults: feasibility and effect on image quality and iodine dose, Radiology, 2004;231: 169–74.
    Crossref | PubMed
  9. Westra SJ, Hurteau J, Galindo A, et al., Cardiac electron-beam CT in children undergoing surgical repair for pulmonary atresia, Radiology, 1999;213:502–12.
    Crossref | PubMed
  10. Lee T, Tsai IC, Fu YC, et al., Using multidetector-row CT in neonates with complex congenital heart disease to replace diagnostic cardiac catheterization for anatomical investigation: initial experiences in technical and clinical feasibility, Pediatr Radiol, 2006;36:1273–82.
    Crossref | PubMed
  11. Westra SJ, Hill JA, Alejos JC, et al., Three-dimensional helical CT of pulmonary arteries in infants and children with congenital heart disease, AJR Am J Roentgenol, 1999;173:109–15.
    Crossref | PubMed
  12. Goo HW, Park IS, Ko JK, et al., Visibility of the origin and proximal course of coronary arteries on non-ECG-gated heart CT in patients with congenital heart disease, Pediatr Radiol, 2005;35:792–8.
    Crossref | PubMed
  13. Abada HT, Larchez C, Daoud B, et al., MDCT of the coronary arteries: feasibility of low-dose CT with ECG-pulsed tube current modulation to reduce radiation dose, AJR Am J Roentgenol, 2006;186:S387–90.
    Crossref | PubMed
  14. Paul JF, Abada HT, Strategies for reduction of radiation dose in cardiac multislice CT, Eur Radiol, 2007;17(8):2028–37.
    Crossref | PubMed
  15. Lambert V, Sigal-Cinqualbre A, Belli E, et al., Preoperative and postoperative evaluation of airways compression in pediatric patients with 3-dimensional multislice computed tomographic scanning: effect on surgical management, J Thorac Cardiovasc Surg, 2005;129:1111–18.
    Crossref | PubMed