Scan time and patient dose for thoracic imaging in neonates and small children using axial volumetric 320-detector row CT compared to helical 64-, 32-, and 16- detector row CT acquisitions

This phantom study has evaluated to what extent acquisition time and patient dose savings can be achieved for thoracic imaging in neonates and small children by using axial volumetric MDCT instead of helical MDCT acquisition. Acquisition protocols were used that yield similar image quality for the different MDCT scanning modes. All measurements were performed on the same scanner that allowed axial as well as helical acquisitions, thereby providing standard conditions where differences found could be ascribed to differences in scanning techniques only. The main findings were that for standard thoracic CT scans of neonates and small children a factor of 5–24 times faster scanning was achieved by volumetric MDCT as compared to helical MDCT. Time saving was most obvious for the largest scan range of 160 mm. At the same time, with volumetric MDCT, dose savings varying between 18% and 40% were achieved a compared to helical MDCT. The smaller the craniocaudal scan range, and the greater the number of detectors rows used for helical scanning (i.e. 64- and 16-MDCT), the greater the percent dose reduction achieved with axial 320-MDCT.

Potential reduction in patient radiation dose by volumetric scanning using a prototype 256-MDCT compared to 16-MDCT has been reported previously for standard 120 kVp applications in adults [13]. In that study, dose saving achieved for chest imaging with a scan range of 38 cm was 28% with axial volumetric 256-MDCT compared to helical 16-MDCT. The dose reduction that can be achieved with axial volumetric scanning compared to helical scanning can be explained by larger beam width of the volumetric MDCT scanner. The actual beam width in a MDCT scanner is the nominal beam width (the product of slice thickness and number of slices) plus a margin that covers the penumbra [13]. The relative contribution of the penumbra becomes less prominent for beams with a larger nominal beam width. Another important aspect that explains patient dose reduction with axial volumetric scanning as,compared to helical scanning is z- over-ranging (or overscanning) at helical acquisitions, which is absent in axial acquisitions. In helical scan mode, the reconstruction algorithm requires additional raw data on both sides of the planned scan range; therefore in helical mode the exposed range exceeds the imaged range. These extra helical rotations at the boundaries outside the planned area contribute to radiation dose, but not to image formation [1]. In axial volumetric acquisitions the exposed range corresponds exactly to the imaged range. Therefore, z-over-ranging is not applicable to volumetric scanning, resulting in a more effective use of radiation for image formation with volumetric acquisition [2].

Although the advantage of volumetric scanning over helical scanning due to lack of z-over-ranging has been well recognized, volumetric scanning has not been applied in neonates or small children for relative large scan ranges (such as chest imaging) with 64-, 32-, or 16-MDCT. While knowledge of the effect of z-over-ranging was obtained with 16-MDCT scanners, performing volumetric acquisitions using these scanners would not be practical for imaging these specific patient groups [1, 2]. This is because only small volumes with a maximum range of 32 mm are, for example, obtained per axial rotation for a 64 MDCT acquisition (by using a Toshiba Aquilion 64-MDCT). Multiple axial rotations would be necessary to image a reasonable scan range. However, multiple rotation acquisitions would hamper “smooth” volume acquisition due to patient motion between these rotations, which is likely to cause step artefacts within the volume. With the 320-MDCT a maximum 160 mm volume is acquired per rotation (simultaneously and without step artefacts) that is well within clinical scan ranges for thoracic imaging in neonates and small children.

Dual-source CT (DSCT) is another recent development and is also used for paediatric radiology [14]. This technology uses two X-ray tubes and two detectors and DSCT particularly improves temporal resolution. Coverage of the volume of interest with DSCT is achieved by a helical acquisition, and requires several rotations. DSCT optimizes the temporal resolution of the scan, but the scan time is longer compared to (axial) volumetric 320-MDCT. Scan time of the DSCT thoracic acquisition described is estimated at the order of magnitude of the 64-slice acquisitions presented in Fig. 1 [14].

In this study the advantage of axial volumetric scanning has been investigated for chest applications only. However, axial volumetric scanning may be applied to numerous clinical situations including abdominal imaging in children and in adults and also for brain, neck, liver, shoulder, hip and knee imaging. In addition, the 320-MDCT scanner allows prospective-gated imaging of the entire heart and coronary arteries within a single rotation, and ventricular functional imaging within a single heart beat.

The reduction in acquisition time and radiation dose that can be achieved with the volumetric MDCT scanner is of particular advantage when imaging neonates and small children. Scan times varying between 1.9 s and 8.3 s as measured here for state-of-the-art helical 64-, 32- and 16-MDCT acquisitions with fast rotation time of 0.35 s, are more prone to image artefacts than the single 0.35-s acquisition time required with volumetric 320-MDCT scanning. The sub-second acquisition time with volumetric scanning may obviate the need for sedation in these children.

Neonates and small children are more susceptible to radiation hazards compared to adults. The sensitivity for stochastic radiation effects (carcinogenesis and genetic effects) of newborns and small children is a factor 10–3 times higher than in adults [3, 15, 16]. Their relatively high susceptibility to radiation and long post-exposure life expectancy requires dedicated paediatric scan protocols [3].

Lowering tube voltage (kVp) and tube current (mA) can be used as effective measures for reducing radiation dose to children [6, 11, 17‐19]. It has been shown that for a given tube current, using a tube voltage of 100 kVp instead of the traditionally used 120 kVp results in a 34% effective dose reduction in children. With 80 kVp instead of using 120 kVp, 68% dose reduction can be achieved [18]. Comparable results were found in another study, where noise level was used as a measure of image quality [6].

At a given tube voltage, by using lower tube current settings in paediatric examinations, effective doses in paediatric chest CT examinations can be lowered by a factor of two or greater as compared to adult examinations [19]. Automatic tube current control, which adapts the acquisition to patient size and shape, effectively contributes to the optimization of CT acquisitions. DLP values are much lower in children than in adults due to the smaller scan lengths, and with proper adaptation of the tube settings, the effective doses in children can be much lower than in adults. Effective patient dose for a standard chest CT is estimated approximately 1.7 mSv in newborns and approximately 5.4–7 mSv in a normal-size adult [17, 20]. Our findings of low DLP in small children, regardless of scan mode, are in line with these estimates.

There are some study limitations that should be addressed. Technology for beam collimation in CT evolves steadily and contributes to the reduction of radiation exposure. Some recent CT scanners from various manufacturers are equipped with a dynamic collimator. The dynamic collimator optimizes the longitudinal dose profile at the start and end of the helical scanned range; it reduces over-ranging, and thus reduces radiation exposure in helical CT scanning [21]. Collimation of axial volumetric CT scanners is being improved by tighter collimation, leading to reduced overbeaming, and thus reduced radiation exposure. The effect of a dynamic collimator in helical scans and the effect of tighter collimation in axial cone beam scans were not investigated in this study. Also, single-rotation axial MDCT imaging results in cone-shaped image borders at the cranial and caudal sides (Fig. 3). Therefore, the diagnostic area available at the peripheral borders of the scanned range is somewhat reduced, although this effect is of minor importance in small objects.

Dose calculations were based on measurement with a single 16-cm dose phantom. Although this might not be fully representative of all patient variations in body shape that may occur in clinical practice, it provides for reasonable dose estimations in children. For craniocaudal scan ranges exceeding 16 cm an extra gantry rotation would be needed; together with table shift this would result in an approximately 1.0–2.0 s extra scan time. However, in our experience, scan ranges exceeding 16 cm are not required for chest imaging in neonates and small children and full chest imaging is performed in 0.35 s rotation. We used DLP to quantify patient dose [4, 10]. Calculation of effective dose in children would allow for calculating risk related parameters and comparison with other radiation-based imaging modalities [17, 22, 23].