Comparison of Different Methods of Calculating CT Radiation Effective Dose in Children Annual Meeting of the Society-for-Pediatric-Radiology Newman, B., Ganguly, A., Kim, J., Robinson, T. AMER ROENTGEN RAY SOC. 2012: W232–W239

Abstract

CT radiation dose is a subject of intense interest and concern, especially in children. Effective dose, a summation of whole-body exposure weighted by specific organ sensitivities, is most often used to compute and compare radiation dose; however, there is little standardization, and there are numerous different methods of calculating effective dose. This study compares five such methods in a group of children undergoing routine chest CT and explores their advantages and pitfalls.Patient data from 120 pediatric chest CT examinations were retrospectively used to calculate effective dose: two scanner dose-length product (DLP) methods using published sets of conversion factors by Shrimpton and Deak, the imaging performance and assessment of CT (ImPact) calculator method, the Alessio online calculator, and the Huda method.The Huda method mean effective dose (4.4 ± 2.2 mSv) and Alessio online calculator (5.2 ± 2.8 mSv) yielded higher mean numbers for effective dose than both DLP calculations (Shrimpton, 3.65 ± 1.8 mSv, and Deak, 3.2 ± 1.5 mSv) as well as the ImPact calculator effective dose (3.4 ± 1.7 mSv). Mean differences ranged from 10.2% ± 10.1% lower to 28% ± 37.3% higher than the Shrimpton method (used as the standard for comparison). Differences were more marked at 120 kVp than at 80 or 100 kVp and varied at different ages. Concordance coefficients relative to the Shrimpton DLP method were Deak DLP, 0.907; Alessio online calculator, 0.735; ImPact calculator, 0.926; and Huda, 0.777.Different methods of computing effective dose for pediatric CT produce varying results. The method used must be clearly described to allay confusion about documenting and communicating dose for archiving as well as comparative research purposes.

View details for DOI 10.2214/AJR.10.5895

View details for Web of Science ID 000306686200012

View details for PubMedID 22826426