Meetings » 4th CTUG Meeting » New Monte Carlo simulations at NRPB for paediatric CT

New Monte Carlo simulations at NRPB for paediatric CT

Paul Shrimpton

Medical Dosimetry Group, NRPB, Chilton, Didcot, Oxon OX11 0RQ

Abstract

Optimisation of protection for paediatric patients undergoing CT is only now beginning to achieve due prominence, following an earlier general lack of awareness of the potentially higher levels of dose to children relative to adults. This unfortunate oversight was fostered, perhaps, by the absence of specific tools for paediatric CT dosimetry. In order to facilitate more comprehensive dose assessment for CT, new Monte Carlo simulations have been performed at NRPB for a complete family of six (MIRD) geometric mathematical phantoms representing ages from newborn to adult. Organ doses have been calculated for CT exposures of contiguous, 1 cm thick, transverse slices in each phantom, and for three CT scanner models (Siemens DRH, GE 9800 and Philips LX) reflecting a range of designs. Calculations have also been carried out in relation to the standard head and body CT dosimetry phantoms. All calculations have been implemented using the MCNP-4C radiation transport code operating on a personal computer (with Pentium III processor).

Rigorous quality assurance procedures have been undertaken so as to validate the models developed. In particular, remarkably good agreement was found with the results of comparable Monte Carlo calculations previously completed at NRPB, in relation to both CT (of adults) and conventional x-ray exposures. Ratios of CTDIw to air kerma were reassuringly close to measured values for the three scanner models.

In general, values of effective dose from CT, when normalised to air kerma, are significantly greater for paediatric patients than for adult patients when using the same technique to scan similar anatomical regions, but with some dependence on the scanner model. The enhancement is greatest for 'head and neck' scans, where it is in the range 2.3-2.6 for the newborn across all three scanners. For scans of the whole trunk it ranges from 1.4 for the scanner without a shaped filter, the Siemens DRH, to 2.3 for the GE 9800 scanner.

The results of the calculations highlight the particular need to use CT scan protocols that are tailored to patient size. A summary of the work has already been published (Khursheed et al, Br J Radiol 75: 819-830 (2002)). The organ dose coefficients will also be published as a software report on the NRPB website and will form the basis for a general method for dose assessment in paediatric CT.

© 2002 NRPB

Back to meeting timetable