This Article Abstract Figures Only Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yates, S J Articles by Goldstone, K E Search for Related Content PubMed PubMed Citation Articles by Yates, S J Articles by Goldstone, K E

Effect of multislice scanners on patient dose from routine CT examinations in East Anglia

East Anglian Regional Radiation Protection Service, Addenbrooke's Hospital, Hills Road, Cambridge CB2 2QQ, UK

	Abstract

Top Abstract Introduction Method Results and discussion Conclusions References

 
As part of the dose optimization process, the Ionising Radiation (Medical Exposure) Regulations 2000 include requirements relating to the assessment of patient dose, and the setting and subsequent review of diagnostic reference levels. In East Anglia, audits of effective dose in CT have been carried out in 1996, 1999 and 2002. In the 2002 audit, nine of the 14 scanners assessed had been replaced since the previous audit. Eight of the new scanners were multislice scanners, acquiring up to 16 slices in a single rotation. The objective of the 2002 audit was to investigate the effect of the introduction of these multislice scanners on patient doses from routine CT examinations. Exposure parameters were collected for 10 different types of routine CT examination. In excess of 550 sets of patient data were obtained. For each of these, effective doses were calculated using the results of Monte Carlo simulations published by the National Radiological Protection Board. Averaged across all 10 examinations, regional mean effective doses are 34% higher than in 1999. The multislice scanners in the region give, on average, 35% more effective dose than the single-slice scanners. The effect of collimation in multislice scanners makes these effective dose differences most notable for examinations that use narrow slice widths. Further optimization of exposures on multislice scanners has the potential to reduce the differences observed between single-slice and multislice doses. However, when taken in combination with the increased use of CT in many hospitals, the effective dose increases observed are likely to result in a significant increase in the already substantial collective radiation dose from CT.

	Introduction

Top Abstract Introduction Method Results and discussion Conclusions References

 
X-ray CT is the technique whereby tomographic images of a patient are obtained from a mathematical reconstruction of X-ray attenuation measurements made through a thin axial slice of the patient. Notwithstanding the undoubted clinical benefits of CT, it is a relatively high dose technique when compared with other imaging modalities. Indeed, whilst CT accounts for only about 3% of all examinations performed using X-rays in the UK, radiation doses from CT account for approximately 40% of the collective radiation dose arising from these medical exposures [1].

In Great Britain, the use of ionizing radiation for medical exposures is subject to the Ionising Radiations Regulations 1999 (IRR99) [2] and the Ionising Radiation (Medical Exposure) Regulations 2000 (IR(ME)R) [3]. With the aim of keeping medical exposures as low as reasonably practicable, IRR99 requires that "such measurements [are made] as are necessary to enable the assessment of representative doses from any radiation equipment to persons undergoing medical exposures" (Reg. 32). Similarly, IR(ME)R requires the employer to establish diagnostic reference levels (DRLs) for standard radiodiagnostic examinations. Employer's procedures should specify that these DRLs "are expected not to be exceeded for standard procedures when good and normal practice regarding diagnostic and technical performance is applied" (Schedule 1).

There have been a number of assessments of patient dose from CT in the past. In 1991 the National Radiological Protection Board (NRPB) published the results of a national survey of CT practice in the UK, which established mean effective doses for a range of examinations [4].

There have however been substantial changes in CT technology since that time. Most recently, the late 1990s saw the introduction into the UK market of scanners with multislice capabilities. These scanners allow the acquisition of several images in a single rotation of the X-ray tube. At a simplistic level, except for having multiple detectors in the axial direction, multislice scanners are physically very similar to single-slice scanners. However, to avoid problems associated with the beam penumbra, it is necessary in multislice CT to irradiate more of the patient than is actually imaged (Figure 1). This effect is of particular significance for narrow slices, where it is estimated that doses could be up to 40% higher than for well-collimated single-slice systems [5].

View larger version (38K): [in this window] [in a new window]   Figure 1. Beam collimation in single-slice and multislice CT. (a) In a well-collimated single-slice system, a 5 mm slice width is achieved by irradiating approximately 5 mm of the single 10 mm wide detector. (b) In a multislice system used to acquire four 1.25 mm slices, the shape of the beam profile means that more than 5 mm of the detector must be irradiated in order to ensure that the four detectors are irradiated to uniform intensity.

At a local level, regional audits of effective dose in CT have been carried out in 1996 and 1999. However, the introduction of multislice technology, combined with significant funding from the National Lottery's New Opportunities Fund, has led to a substantial change in regional CT equipment since 1999, as is shown in Table 1.

	Method

Top Abstract Introduction Method Results and discussion Conclusions References

 
Previous work has shown that, when auditing patient doses in CT, standard protocols can be of limited use in assessing actual patient dose [7]. For this reason actual patient exposure data was sought from each scanner, for a minimum of 10 patients for each of 10 common categories of CT examination (head, neck, routine chest, high-resolution chest, chest-abdomen, chest-abdomen-pelvis, abdomen-pelvis, abdomen, pancreas and lumbar spine). Audit forms were distributed to CT superintendent radiographers at each hospital, requesting information relating to the exposures carried out, as well as information relating to patient gender, height and weight. Following recommendations made for patient dose measurements in radiography [8], patients with weights outside the range 70±20 kg were excluded from the final analysis.

Effective doses were calculated using the results of Monte Carlo simulations carried out in the early 1990s by the NRPB [9]. These results allow equivalent doses to the major organs in the body to be calculated for the irradiation of a specified 5 mm wide slab of an anthropomorphic phantom. Each organ dose is calculated as the product of the X-ray tube current, the tube rotation time, the normalized CT dose index measured free-in-air (_nCTDI_100,air) for the beam collimation, tube potential and beam filtration used, and the appropriate normalized organ dose provided by the NRPB. For a complete acquisition, consisting of the irradiation of several slabs of the phantom, organ doses resulting from the irradiation of each slab are summed, and a final effective dose is calculated by applying tissue weighting factors according to ICRP60 [10].

Data analysis was carried out using in-house spreadsheets linked to the ImPACT CT patient dosimetry calculator [11], which provides a regularly updated convenient user interface for the NRPB's data. The ImPACT dosimetry calculator also identifies which of the NRPB data-sets is most appropriate for use with modern scanners that did not exist at the time of the original Monte Carlo simulations. This scanner matching is based on the concept of the ImPACT factor, which is related to the effective beam energy, and has been shown to correlate well with the effective dose from a scanner [12]. The ImPACT dosimetry calculator also provides generic values of _nCTDI_100,air for a range of scanner models. However, in this work, measured values of _nCTDI_100,air were used in the dose calculation procedure. These results were collected as a part of routine quality control measurements, using a 100 mm pencil ionization chamber with an air-kerma calibration traceable to national standards via a therapy-level secondary standard.

Translating an individual patient exposure onto the anthropomorphic phantom is relatively straightforward in most situations. However, where a patient's height is significantly different from the size of the phantom, use of the actual scanned length can result in the irradiation of too few, or too many, of the organs in the mathematical phantom. To overcome this potential discrepancy, our audit forms included a diagram of the human skeleton, on which radiographers reported the upper and lower extent of each scan. This information was used in the selection of the part of the phantom irradiated, improving the correspondence between the organs irradiated in the patient and the phantom.

Accurately modelling head examinations using the Monte Carlo data is also hindered because head scans are often performed with a tilted gantry, whereas the NRPB data only allow for the calculation of doses from slices acquired perpendicular to the long axis of the patient. Whilst the effect of gantry tilt on the equivalent dose to the lens of the eye will be appreciable, the effect of tilt on the effective dose will be much less significant. These scans were therefore modelled using standard slices, perpendicular to the long axis of the patient, with the selection of slices being made so as to best represent the scan actually performed.

In several cases the scanners assessed were capable of some form of automatic dose optimization. This is normally achieved by modulating the mAs during a spiral acquisition, based either on a previously acquired scan projection radiograph (SPR), or on the preceding rotation in the spiral acquisition. Such facilities make dose calculations more complicated, because a single mAs cannot be assumed for the whole acquisition. Where appropriate, participating staff were therefore asked to report acquisitions in sections, with each section representing a region of the body where the mAs remained reasonably constant. Typical mAs values were reported for each section, based on the mAs values reported on each image. Where available, other information, such as maximum mAs, minimum mAs, total mAs and dose–length product (DLP) were also reported. These values allowed for some verification of the estimates of "typical mAs" made.

In addition to calculations of effective dose, DLPs were also calculated for each procedure. The DLP is defined as the pitch corrected weighted CT dose index (CTDI_w) multiplied by the scanned length. The CTDI_w is an estimate of the average dose to a standard CT dosimetry phantom from a single axial CT slice. It is calculated as the weighted average of CTDI measurements made at the centre (CTDI_100,c) and periphery (CTDI_100,p) of a 16 cm diameter (head) or 32 cm diameter (body) polymethylmethacrylate phantom, according to Equation (1).

Whilst modern scanners normally report DLP for examinations, routine quality control measurements on scanners in the region have shown that discrepancies often exist between reported and measured values. Calculated DLPs, based on measured values of CTDI_w, were therefore used in this audit, avoiding the issue of these discrepancies, and allowing older scanners not reporting DLP to be included in the analysis. CTDI_w measurements were made on each scanner using the pencil ionization chamber and standard CT dosimetry phantoms described previously.

	Results and discussion

Top Abstract Introduction Method Results and discussion Conclusions References

 
Audit forms were returned for a total of over 550 patient examinations from the 12 scanners assessed. (Although there were 14 scanners in the region in 2002, at 2 hospitals there were pairs of identical scanners. No attempt was made to distinguish between the two scanners at each of these hospitals, as in both cases the two scanners were set up and used identically.)

Effective doses
Table 2 summarizes the overall results of the 2002 audit, and provides results from the 1999 audit [13] for comparison. These results show that on average, mean effective doses from CT in the region are 34% higher in 2002 than in 1999.

The relative effective doses for each scanner were then averaged across all examination categories, to give a measure of the overall performance of each scanner (Figure 3).

View larger version (52K): [in this window] [in a new window]   Figure 2. Mean effective doses for each scanner, for each examination category. Scanners are anonymously identified by the letters A to L, according to the order in which they appear in Figure 3. Varying numbers of scanners contributed data to the different examination categories. Error bars represent the standard error of each mean.

View larger version (45K): [in this window] [in a new window]   Figure 3. Mean relative effective doses (averaged across all examinations) for each scanner in the 2002 audit. Error bars represent the standard error of each mean.

The single-slice scanner with apparently high effective doses (scanner I) is one of the oldest single-slice scanners in the audit. This scanner is restricted to operate at 130 kV and compared with other scanners, has a relatively small amount of beam filtration. Consequently values of _nCTDI_100,air, i.e. tube output in mGy/mAs, are relatively high for this scanner. It is noted however that the mAs values used on this scanner are very similar to those used for similar examinations on other single-slice scanners. The result is relatively high patient doses, as observed.

The average relative effective dose for all single-slice scanners is 0.85±0.10 (mean±standard error. The uncertainty quoted here is representative of the variation between scanners; the uncertainty in the underlying Monte Carlo simulation and effective dose calculation has not been quantified in this study). Similarly, the average relative effective dose for all multislice scanners is 1.15±0.06. The multislice scanners in the region therefore give, on average, 35% more effective dose than the single-slice scanners. This difference is not seen uniformly across all examinations however. As demonstrated in Figure 2, the distinction between single-slice and multislice effective doses is generally seen to be greatest for examinations using narrow slices, e.g. head, high resolution chest, but is less apparent for other examinations, e.g. abdomen-pelvis. This can again be explained by reference to the beam collimation effect in multislice CT. Assuming that it is necessary to irradiate a constant amount more of the patient than is actually used for imaging, the geometric efficiency (the ratio of total imaged width to irradiated width) will be poorest where a small number of narrow slices are acquired in a single rotation. For example, irradiating 7 mm of patient to acquire four 1.25 mm slices results in a geometric efficiency of about 70%, whereas irradiating 22 mm of patient to acquire four 5 mm slices maintains a geometric efficiency of about 90%.

It is of note that, in the majority of cases where scanners have remained unchanged since 1999, the effective doses assessed in 2002 were very similar to those in 1999. On the assumption that clinical practice has not changed significantly, this provides confidence in the reliability of the assessment method. For example, for scanner C, 2002 effective doses were within ±10% of 1999 values for seven of the 10 examination types. Discrepancies were found to be greater in situations where the number of cases returned in either 1999 or 2002 was small, and so there are correspondingly larger uncertainties in the quoted values.

Dose–length products and diagnostic reference levels
In addition to effective doses, DLPs were calculated for each examination in the 2002 audit. Results are summarized in Table 3. In the majority of cases, relatively high effective doses at individual hospitals are predictably reflected by relatively high DLPs, and vice versa. The value of including DLP results, however, is that modern scanners normally quote the DLP for an examination, and as such it is the most useful quantity for a DRL. Working Parties on DRLs have recommended a clear distinction between Local and National DRLs [14, 15]. Separate Local DRLs should be set for each piece of equipment, and may be calculated as the mean "dose" received by a set of standard patients on that equipment. As such, the mean DLPs established for each examination for each scanner in this audit represent appropriate Local DRLs for each scanner.

Comparison with European and National data
Whilst there are no formal National DRLs with which to compare our results, comparison can be made with European reference levels recommended in 1999 [6], and mean effective doses from the NRPB's dose audit in 1991 [4]. These data are summarized in Table 4. DLPs all comply with European Reference Levels for routine chest, high resolution chest and abdomen examinations. There are however two scanners for which the mean head DLP exceeds the European reference value of 1050 mGy.cm. These European reference values pre-date multislice CT however, and exceeding this value is perhaps further evidence of the need for more up-to-date National and European CT dose information. Again, it should be noted that head examinations are one of those examinations where significant effective dose differences have been observed between single-slice and multislice scanners.

Dose optimization
In this audit we have focused solely on CT doses, and have made no attempt to compare image quality, or the diagnostic value of images, between scanners. It is of note that at the time of this audit many of the multislice scanners had been in clinical use for less than a year. In many cases examination protocols were therefore still very strongly influenced by settings recommended by the manufacturer, as there had been insufficient time for major optimization of protocols.

Since the data for this audit were collected, one of the highest dose multislice scanners (scanner K) has received a software upgrade, and is now capable of automatic tube current modulation. Initial results from the introduction of this feature into clinical use suggest that tube currents are typically 20% lower than those used in standard protocols. The introduction of this and further dose reduction technology, together with smaller scale improvements to individual examination protocols, will reduce some of the doses reported in this audit. It is therefore predicted that, in time, some of the distinction between single-slice and multislice doses reported here will be lost, as successive improvements reduce doses from multislice scanners.

The importance of dose optimization is highlighted further if, rather than individual doses, collective doses from CT are considered. No attempt has been made here to assess changes in examination frequencies in the period 1999 to 2002. A simplistic look at the number of scanners in the region (10 in 1999, 14 at the time of the audit in 2002, and 16 currently) suggests however that there has been a significant increase in the use of CT over this time period. Combined with the higher doses observed in this audit for multislice scanners, there is clearly the potential for a significant increase in the collective radiation dose received by patients undergoing CT examinations. It is predicted that in future reviews of medical exposures of the UK population, the percentage contribution from CT is likely to increase beyond the 40% at which it currently stands.

	Conclusions

Top Abstract Introduction Method Results and discussion Conclusions References

 
Effective doses from 10 routine CT examinations have been assessed for 14 CT scanners in the East Anglian region, using data from actual patient exposures. Averaged across all examinations, effective doses were found to be 34% greater than when a similar audit was carried out in 1999. In the intervening period, eight multislice scanners have been introduced into the region, and in the 2002 audit the multislice scanners were found to give, on average, 35% more effective dose than the single-slice scanners. However, at the time of the audit, there had been relatively little opportunity for the detailed optimization of exposure protocols on many of the multislice scanners. It is therefore anticipated that further optimization of protocols, together with an increase in the availability and use of automatic tube current modulation, will help to reduce doses from multislice CT in the future. Despite this, when combined with the increasing number of CT scanners across the region, and in the UK as a whole, it would appear that the collective radiation dose from CT to the UK population is likely to continue to rise beyond current estimates. As would be expected from a simplistic consideration of geometric efficiencies, the distinction between single-slice and multislice effective doses was found to be greatest for examinations using narrow slice widths, where the effect of beam collimation in multislice CT is most severe. Given the differences observed between single-slice and multislice scanners, it is suggested that separate National DRLs might need to be set for single-slice and multislice CT. As a minimum however, National DRLs will need to be set with sufficient flexibility to allow for the potentially higher doses from multislice CT.

	Acknowledgments

 
We are grateful to Mr M J White and Dr J P Eatough for their work in carrying out the 1999 dose audit. We would also like to thank all the radiographers at hospitals in the region who contributed data toward the 2002 audit.

	Footnotes

 
This work was funded in part by Access to Learning for the Public Health Agenda (ALPHA), formerly known as the Anglia Clinical Audit and Effectiveness Team (ACET).

Received for publication July 22, 2003. Revision received November 4, 2003. Accepted for publication November 12, 2003.

	References

Top Abstract Introduction Method Results and discussion Conclusions References

 

1. Hart D, Wall BF. Radiation exposure of the UK population from medical and dental X-ray examinations (NRPB-W4). Chilton: National Radiological Protection Board, 2002.
2. The Ionising Radiations Regulations 1999 (SI 1999/3232). London: The Stationery Office, 1999.
3. The Ionising Radiation (Medical Exposure) Regulations 2000. London: The Stationery Office, 2000.
4. Shrimpton PC, Jones DG, Hillier MC, Wall BF, Le Heron JC, Faulkner K. Survey of CT practice in the UK. Part 2: dosimetric aspects (NRPB-R249). Chilton: National Radiological Protection Board, 1991.
5. Lewis MA. Multislice CT: opportunities and challenges. Br J Radiol 2001;74:779–81.[Free Full Text]
6. Bongartz G, Golding SJ, Jurik A, Leonardi M, van Meerten EP, Geleijns J, et al. European guidelines on quality criteria for computed tomography (EUR 16262 EN). Luxembourg: European Commission, 1999.
7. Wade JP, Weyman JC, Goldstone KE. CT standard protocols are of limited value in assessing actual patient dose. Br J Radiol 1997;70:1146–51.[Abstract]
8. Holubinka MR, Jones AP, Rawlings DJ, Roberts PJ, Robertson J, Wall BF. National protocol for patient dose measurements in diagnostic radiology. Chilton: National Radiological Protection Board, 1992.
9. Jones DG, Shrimpton PC. Normalised organ doses for X-ray computed tomography calculated using Monte Carlo techniques (NRPB-SR250). Chilton: National Radiological Protection Board, 1993.
10. 1990 Recommendations of the International Commission on Radiological Protection (ICRP 60). Oxford: Pergamon Press, 1991.
11. Keat N. ImPACT CT patient dosimetry calculator (version 0.99 m). London: ImPACT, 2002. (www.impactscan.org/ctdosimetry.htm)
12. Lewis MA, Edyvean S, Sassi SA, Kiremidjian H, Keat N, Britten AJ. Estimating patient dose on current CT scanners: Results of the ImPACT CT dose survey. London: ImPACT, 1999. (www.impactscan.org/dosesurveysummary.htm).
13. White MJ, Eatough JP, Goldstone KE. Audit of effective dose for patients undergoing X-ray computed tomography. Cambridge: East Anglian Regional Radiation Protection Service, 1999.
14. Workman A, Kotre J, Shaw A, Fong R, Wall B, Bury R, et al. IPEM/NRPB/RCR/CoR/BIR diagnostic reference levels working party. IPEM Newsletter 2000;67:2–4.
15. Workman A. Reference doses and DRLs: the background and the publication. In: Proceedings of UK Radiological Congress 2003; 2003 June 15–17; Birmingham. London: British Institute of Radiology, 2003.

This article has been cited by other articles:

				R. S. Livingstone and P. M. Dinakaran Regional survey of CT dose indices in India Radiat Prot Dosimetry, September 1, 2009; 136(3): 222 - 227. [Abstract] [Full Text] [PDF]

				J Rixe, G Conradi, A Rolf, A Schmermund, A Magedanz, D Erkapic, A Deetjen, C W Hamm, and T Dill Radiation dose exposure of computed tomography coronary angiography: comparison of dual-source, 16-slice and 64-slice CT Heart, August 15, 2009; 95(16): 1337 - 1342. [Abstract] [Full Text] [PDF]

				O J ARTHURS, S J YATES, P A K SET, D A GIBBONS, and A K DIXON Evaluation of image quality and radiation dose in adolescent thoracic imaging: 64-slice is preferable to 16-slice multislice CT Br. J. Radiol., February 1, 2009; 82(974): 157 - 161. [Abstract] [Full Text] [PDF]

				J A SOYE and A PATERSON A survey of awareness of radiation dose among health professionals in Northern Ireland Br. J. Radiol., September 1, 2008; 81(969): 725 - 729. [Abstract] [Full Text] [PDF]

				R. M. Kwee and T. C. Kwee Imaging in Local Staging of Gastric Cancer: A Systematic Review J. Clin. Oncol., May 20, 2007; 25(15): 2107 - 2116. [Abstract] [Full Text] [PDF]

				P C Shrimpton, M C Hillier, M A Lewis, and M Dunn National survey of doses from CT in the UK: 2003 Br. J. Radiol., December 1, 2006; 79(948): 968 - 980. [Abstract] [Full Text] [PDF]

				A. N. Primak, C. H. McCollough, M. R. Bruesewitz, J. Zhang, and J. G. Fletcher Relationship between Noise, Dose, and Pitch in Cardiac Multi-Detector Row CT RadioGraphics, November 1, 2006; 26(6): 1785 - 1794. [Abstract] [Full Text] [PDF]

				M A Lewis and S Edyvean Patient dose reduction in CT Br. J. Radiol., October 1, 2005; 78(934): 880 - 883. [Full Text] [PDF]

				S. J. Golding Multi-slice computed tomography (MSCT): the dose challenge of the new revolution Radiat Prot Dosimetry, May 17, 2005; 114(1-3): 303 - 307. [Abstract] [Full Text] [PDF]

				BJR Review of the Year - 2004 Br. J. Radiol., March 1, 2005; 78(927): 181 - 185. [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yates, S J Articles by Goldstone, K E Search for Related Content PubMed PubMed Citation Articles by Yates, S J Articles by Goldstone, K E