Development of a generalized method to allow the estimation of doses to the ICRP reference adults from CT, on the basis of normalized organ and CTDI dose data determined by Monte Carlo calculation for a range of contemporary scanners. (7th February 2023)

• Record Type: Journal Article
• Authors: Jansen, Jan TM; Shrimpton, Paul C; Edyvean, Sue
• Is Part Of: Physics in medicine & biology. Volume 68:Number 3(2023)
• Publication Date: 2023-02-07
• Subjects: computed tomography -- Monte Carlo radiation transport code -- organ dose coefficients -- effective dose coefficients -- match CT scanners -- adult reference computational phantoms; Biophysics -- Periodicals; Medical physics -- Periodicals; 610.153
• DOI: 10.1088/1361-6560/acb2a8
• Languages: English
• ISSNs: 0031-9155

Abstract

Abstract: Objective . Development of a method to provide organ and effective dose coefficients to reference adults for any CT scanner based on values of CTDI measured both in air and in standard CT dosimetry phantoms. Approach . Results from previous Monte Carlo simulations for a range of contemporary CT scanners have been analyzed to provide linear models relating values of organ dose (normalized to CTDI free-in-air ), for each slab of 3 reference phantoms (ICRP Male/Female, and AH hermaphrodite), to similarly normalized values of CTDI in standard CT dosimetry phantoms. Three methods have been investigated to apply the models to values of CTDI for a 'new' scanner not previously simulated: a Generic approach using averaged normalized organ dose profiles for whole body exposure of the phantoms; and two processes for matching the scanner, on the basis of normalized organ doses or effective dose ( n E 103, phan ), to one of the 102 sets of dose coefficients previously calculated for 12 contemporary CT scanner models, from 4 manufacturers, operating under a range of conditions. Main results . The merit of each method has been quantitatively assessed when applied to both the present contemporary scanners with each test data set being excluded in turn during the matching process, and also to 3 previously-simulated older scanners. Whereas all three methods appear viable, with all doses being within 1% and 10% for the contemporary and old scanners respectively, matching to n E 103, phan is overall the approach preferred in practice, yielding an uncertainty of around 6% in estimated values of n E 103, phan . The present methodology also provides superior performance when compared against some other common normalization factors for E 103, phan . Significance . The CT dose model and the data sets will be incorporated into a new CT dosimetry tool that will be made available from UKHSA in support of facilitating improvements in patient protection.